Effects of elevated CO2 levels on lung immune response to organic dust and lipopolysaccaride

Workplaces with elevated organic dust levels such as animal feed barns also commonly have elevated levels of gasses, such as CO2. Workers exposed to such complex environments often experience respiratory effects that may be due to a combination of respirable factors. We examined the effects of CO2 at the ASHRAE recommended limit (1000 ppm) as well as the EPA 8hr time weighted average limit (5000 ppm) on lung innate immune responses in mice with exposure to inammatory lipopolysaccharide and organic dust. Mice were nasally instilled with dust extracts or LPS and immediately put into chambers with a constant ow of room air (approx. 430 ppm CO2), 1000 ppm, or 5000 ppm CO2 enriched air. Organic dust exposures tended to show decreased inammatory responses with 1000 ppm CO2 and increased responses at 5000 ppm CO2. Conversely, LPS with addition of CO2 as low as 1000 ppm tended to inhibit several inammatory markers. In most cases saline treated animals showed few changes with CO2 exposure, though some changes in mRNA levels were present. This shows that CO2 as low as 1000 ppm CO2 was capable of altering innate immune responses to both LPS and organic dust extracts, but each response was altered in a different fashion.


Abstract
Workplaces with elevated organic dust levels such as animal feed barns also commonly have elevated levels of gasses, such as CO2. Workers exposed to such complex environments often experience respiratory effects that may be due to a combination of respirable factors. We examined the effects of CO2 at the ASHRAE recommended limit (1000 ppm) as well as the EPA 8hr time weighted average limit (5000 ppm) on lung innate immune responses in mice with exposure to in ammatory lipopolysaccharide and organic dust. Mice were nasally instilled with dust extracts or LPS and immediately put into chambers with a constant ow of room air (approx. 430 ppm CO2), 1000 ppm, or 5000 ppm CO2 enriched air. Organic dust exposures tended to show decreased in ammatory responses with 1000 ppm CO2 and increased responses at 5000 ppm CO2. Conversely, LPS with addition of CO2 as low as 1000 ppm tended to inhibit several in ammatory markers. In most cases saline treated animals showed few changes with CO2 exposure, though some changes in mRNA levels were present. This shows that CO2 as low as 1000 ppm CO2 was capable of altering innate immune responses to both LPS and organic dust extracts, but each response was altered in a different fashion.

Background
Organic dust exposures in animal feed operations has long been shown to be detrimental to the health of workers in these facilities 1,2 . Experimental approaches to determining the factors present in these dusts that may be responsible for respiratory in ammation have yielded considerable data, particularly on the role of microbial products such as endotoxins (LPS) and proteoglycans, and others as potential causes of these lung problems 3,4 . Such dust exposures however are often in the context of facilities with a variety of elevated gases such as ammonia and hydrogen sul de, both of which have been implicated in causing lung problems or being immuno-modulatory 5,6 .
Carbon dioxide (CO2) has often been overlooked despite signi cant data showing routine elevations of CO2 in animal feed operations 7 often as high as the OSHA 8hr time weighted average (TWA) limit of 5000 ppm,. CO2 at levels as low as 1000 ppm have been shown to induce cognitive changes in humans 8,9 . Indeed, elevated CO2 is a broader problem, and common in many facilities such as schools 9 , daycares 10,11 , prisons 12 , cars 13 , airplanes 14 , and many other locations 15 .
Previously we have shown that innate immune responses to organic dust extracts were changed when co-exposed to CO2 at 5000 ppm 16 . These changes were at the protein and mRNA levels and in most cases were indicative of enhanced in ammation. Subsequent to this study others showed that neutrophils were altered by similar levels of CO2 17 . At present however these are the only studies we are aware of to address this issue in lung immunology. There have been several studies related to behavior 18 and at higher levels of CO2 where there is induction of hypercapnic acidosis in animals and cell cultures [19][20][21] . Results for these latter studies have been somewhat mixed, particularly in situations of LPS exposure versus more complex challenges such as tobacco extracts 21,22 . These studies also do not address CO2 at levels applicable to most work environments and crowded spaces 23,24 .
The aims of the current study were to extend the previous co-exposure model to assess lower levels of CO2 exposure while at the same time testing both a de ned single in ammatory stimuli (LPS) and a complex one (organic dust extracts). We hypothesize that the lung alters innate immune responses when presented with a combined exposure of an in ammatory agent and CO2 as low as 1000 ppm.

Organic Dust Extracts and Treatments
Organic dust extracts (HDE) were produced from settled dust samples collected from two separate swine con nement facilities in Nebraska. These dust extracts have been characterized previously for muramic acid, endotoxin, and protein 25 , and the bacterial composition has also been described 26 . Extracts were prepared by mixing 1 g dust with 10 ml HBSS (without calcium, Sigma, St. Louis, MS) and incubating for 1hr at room temperature before centrifugation twice, and subsequent lter sterilization through a Nalgene 0.2uM SFCA membrane syringe lter (ThermoFisher, Rochester, NY) as described previously 27 . This results in a solution of approximately 0.105 g/ml dust that is stable at -20C or lower for at least a year or more. Dust extracts were given to mice at a nal concentration of 12.5% vol/vol or 0.005 g/ml dust delivered in 40 ul dropwise to the nares, and mice allowed to inhale under light (1%) iso urane anesthesia.
LPS from Escherichia coli 0111:B4 (lot 095M4163V, Sigma, St. Louis) was diluted in HyClone HBSS(1X) (HyClone Laboratories; Logan, UT) to 0.1 µg, and delivered in 40 ul to the nares, and mice allowed to inhale under light (1%) iso urane anesthesia. Control groups were given 40 ul HBSS in the same manner under light anesthesia.

Animals
Six to seven week old female C57BL/6 mice were purchased from Charles River Laboratories (Montreal, Quebec, Canada). Mice were housed in standard cages in a controlled room with ambient temperature of 23 o C +/-2 o C and a 12-hour light/dark cycle. Mice were fed standard laboratory chow and water ad libitum. Mice were acclimatized in their rst week in facility for 2 hours per day for 5 days to plethysmography chambers (part 601-1425-001; DSI, Minneapolis, MN) ventilated by a Buxco Finepointe whole body plethysmography 4-site system (DSI, Minneapolis Minnesota). All intranasal treatments were given between 8:00-8:30AM to minimize any potential circadian rhythm differences between treatments. At end of the experiment mice were euthanized.

CO2 Delivery and Whole Body Plethysmography
All mice upon intranasal treatment were placed immediately into whole body plethysmography (WBP) chambers (DSI, Minneapolis, MN) for the duration of exposure under positive air ow of 5psi. The bias ow air consisted of either of room air (~ 500 ppm CO2), or air from tanks consisting of room air enriched for CO2 to 1000, or 5000 ppm CO2 (Praxair; Mississauga, Canada). Air ow was constant through the exposure period to prevent any accumulation of CO2 due to respiration from animals. Mice were provided with food and water in exposure chambers for the duration of exposures. CO2 readings of the treatment room provided an average ambient CO2 level of 430 ppm averaged over 8 using a Q-Trak Indoor Air Quality Monitor 7575 (TSI; Shoreview, MN).
Readings from the WBP chambers were captured every 2 seconds for the entire duration of exposure.
Results were averaged for 20 minute intervals using the Finepointe system software (DSI, Minneapolis, MN).

Bronchoalveolar Lavage
Lungs were lavaged as detailed earlier (Romberger et al. 2002). Brie y, lungs were washed three times with 0.5 ml HBSS each time. Bronchioalveolar lavage (BAL) uid was centrifuged at 1000 X g for 10 minutes, and supernatant transferred to new tubes and stored at -80oC. Cell pellets were resuspended in HBSS and stained with trypan blue (Life Technologies, Grand Island, NY) and counted using a hemocytometer. Cells were resuspended to 100 ul in HBSS and adhered to glass slides (Fisher Scienti c; Pittsburgh, PA) via cytocentrifugation for 10 min at 10000 X g (Cytopin, Shandon Elliott, Great Britain). Cells were dried overnight and xed and stained using Diff Quik kit (Siemens Healthcare Disagnostics, Newark, DE) and mounted using MM24 mounting media (Leica Biosystems; Buffalo Grove, IL). A differential count of 100 cells was made based on morphological assessment of cells and expressed as absolute cell numbers.

Lung collection
After BAL collection lungs were excised. The right lung was tied off at the primary bronchus, removed, ash frozen in liquid nitrogen, and stored at -80oC. The left lung was slowly in ated with 200 ul of 4% PFA (Sigma; St. Louis, MO) and stored in 4% PFA overnight, followed by 100% ethanol. The xed lung was embedded the next day in para n in a Intelsint RVG/1 tissue processor (Intelsint; Turin, Italy) followed by mounting in para n blocks in Tissue Tek II tissue embedder (Sakura Finetek; Nagano, Japan).

Lung Histology
Lung sections from para n embedded lungs were sectioned in 5 uM slices on a Microm 350S microtome (Microm, Germany). Tissue sections were then stained with Hemotoxylin and Eosin as previously described (Rolls and Sampias 2019) and mounted using Surgipath MM24 (Leica Biosystems, Richmond, IL). Sections were imaged using an Aperio CS2 virutual microscopy system (Leica Biosystems, Concord ON Canada) and examined for any notable changes to lung structure. ELISA Quanti cation of MCP-1, MIP-2, IL-6, KC, IL-1b, TNF-a, IL-4, IL-5, IL-13, IL-33, and IL-10 in BAL uid was determined by Luminex xMAP multiplex ELISA (Procartaplex, ThermoFisher) according to manufacturer's speci cations. Samples were read using a Bioplex 200 system (Bio-Rad, Hercules, California) and Bioplex Manager Software (Bio-Rad) RNA puri cation and RT-PCR analysis RNA was puri ed from lung tissue samples using a Qiagen RNeasy Plus Mini kit (Qiagen, Chatsworth CA) according to the manufacturer's instructions for tissue extraction. Brie y, 20-25 mg of lung tissue was put into 350 ul of RLT lysis buffer and 2.0 mm zircon beads (BioSpec, Bartlesville, OK). Samples were placed in a Biospec Mini-BeadBeater-24 (BioSpec) and run twice for 2 minutes, with a pause between runs of 5 minutes where samples were held on ice to prevent heat buildup. Supernatant was removed to DNA eliminator column, and the remainder of puri cation done according to the kit protocol. RNA was quanti ed by Take3 plate (Biotek, Winooski, VT) in a Synergy HT plate reader (BioTek). cDNA synthesis was done using the iScript Reverse Transcription Supermix for RT-qPCR (Bio-Rad, Hercules, CA) with 300 ng of template mRNA. Samples were incubated at 25oC for 5 min, 46oC for 20 minutes, and nally at 95oC for 1 minute in a Bio-Rad CFX96 Touch Real-Time PCR Detection System (BioRad). RT-PCR was done using probes for TLR2 (Mm00442346_m1), TLR4 (Mm445273_m1), Hsp72 (Mm01159846_S1), A20/TNFAIP3 (Mm00437121_m1) (Life Technologies, Grand Island, NY). Ribosomal RNA (Life Technologies) was used as an endogenous control in all reactions. PCR was conducted using a Bio-Rad CFX96 Touch Real-Time PCR Detection System (Bio-Rad). Reactions were all carried out in duplicate and performed 2 minutes at 50oC, 10 minutes at 95oC, then 40 cycles of 15 seconds at 95oC and 1 at 60oC minute each using ddPCR Supermix for Probes kit (Bio-Rad). Relative comparison of selected targets to the ribosomal endogenous control cycle threshold (CT) value was analyzed with the DDCt method.

Statistical Analysis
Data was analyzed using GraphPad Prism 6 (GraphPad Software, San Diego, CA). Error bars represent mean +/-SEM. For values outside the assay limit of detection either the LOD/2 or a minimum value below the lowest attained value was designated. Statistical signi cance was determined using one-way ANOVA with follow-up Tukey test for multiple comparisons. If the assumption of equal variance was not met, the data was either log transformed followed by one-way ANOVA and multiple comparison tests, or nonparametric tests (Kruskal-Wallace) were conducted. For all tests, a p-value < = 0.05 was considered signi cant for differences between groups.

Plethysmography Data
Monitoring of breathing parameters of mice in exposure chambers showed changes in several parameters over the six hours of exposure, across all treatments (Fig. 1).
Comparing within and between treatments there was increased tidal volume with all co-treatments with CO2 compared to treatments with no CO2.
None of the treatments appeared to have any signi cant effect on breathing frequency (Fig. 1b), total inspiratory time (Fig. 1f), or minute volume ( Fig. 1 g).

Bronchoalveolar Lavage Cells
Cell counts from BAL uid showed several patterns (Fig. 2). First, mice given saline intranasally showed no signi cant changes in BAL cell composition or number with increased CO2 (Fig. 2a).
LPS treatment showed an increase in macrophages/monocytes compared to saline control. This increase was signi cant or near signi cant compared to treatment at a similar CO2 level (Saline vs. LPS (p = 0.0637), Saline vs. LPS + 1000 ppm CO2 each (p = 0.0852), Saline vs. LPS + 5000 ppm CO2 each (p < 0.05). No signi cant increases in neutrophils were noted between saline and LPS treatments, except between Saline + 5000 ppm CO2 and LPS + 5000 ppm CO2 (p < 0.01).
HDE exposure (Fig. 2c) on its own or with room air was not able to induce a signi cant rise in monocytes/macrophages. However co-exposure of HDE with 1000 ppm CO2 was able to induce an increased numbers of cells over saline given at a similar level of CO2 (p = 0.0732), which was signi cantly increased with 5000 ppm co-exposure (p < 0.05). Levels of neutrophils were elevated with HDE, but not signi cantly different as compared to saline. However, co-exposure of HDE with 5000 ppm CO2 caused a substantial increase of the neutrophil population as compared to HDE and room air (p < 0.001), which was also signi cantly different from the HDE and 1000 ppm CO2 co-exposure (p < 0.01), and from the saline and 5000 ppm CO2 co-exposure (p < 0.0001).

ELISAs
As shown in Figs. 3 − 1 and 3 − 2, the expression of cytokines appeared to follow three general patterns.
First, no difference in cytokines was noted when CO2 was co-exposed with saline.
For LPS, increased CO2 provoked a decrease in cytokines/chemokines as compared to co-exposure with room air. Co-delivery of 1000 ppm CO2 with LPS resulted in signi cant decreases in IL-6 (p < 0.01), KC (p < 0.001), TNF-a (p < 0.01), and IL-1b (p < 0.01) as compared to co-exposure with room air. Similar trends were noted at 5000 ppm CO2 co-exposure with LPS with reductions in IL-6 (p < 0.05), KC (p < 0.001), and TNF-a (p < 0.01) as compared to room air co-exposure to LPS. Cytokine/chemokines were not signi cantly different between co-exposure of LPS to CO2 at 1000 ppm and 5000 ppm as compared to saline co-exposures at these same CO2 levels.
The situation was more complex with HDE. Compared to co-exposure with room air, at 1000 ppm CO2 coexposure HDE showed no signi cant differences between cytokines/chemokines though there was some apparent drop in production. However at 5000 ppm CO2 co-exposure there was a signi cant rise in MCP-1 and MIP-2 (p < 0.05) compared to HDE on room air. There was also a raise in MCP-1 (p < 0.001), MIP-2 (p < 0.01), and IL-1b (P < 0.01) at 5000 ppm codelivery as compared to co-delivery of 1000 ppm CO2. This suggested an increase in in ammatory response at higher CO2 delivery as opposed to the reductions seen with all CO2 co-delivery levels when CO2 was co-exposed with LPS. Further, with HDE co-exposure with 1000 ppm CO2 there was a tendency to reduced cytokines whereas 5000 ppm CO2 co-delivery with HDE showed increases in cytokines, suggesting differential responses with increased CO2 concentrations when co-exposed with HDE.

RT-PCR
We looked for evidence of changes to signaling in lung cells rst at the receptor level with two of the receptors most commonly associated with response to HDE, TLR2 and TLR4, the latter of which is critical to responses to LPS. TLR2 (Fig. 4A), which is implicated in HDE induced in ammation 28,29 showed no signi cant change with HDE co-exposure with any level of CO2. However, LPS with room air induced a signi cantly large increase in TLR2 mRNA compared to saline and room air (p < 0.05), and this increase was inhibited by addition of 1000 ppm CO2 (p < 0.01) and 5000 ppm CO2 (NS) co-exposures. TLR4 (Fig. 4B) showed no signi cant change in mRNA expression across all LPS and HDE treatments and CO2 co-exposures.
With little work having been done on CO2 signaling pathways we next looked at a negative regulator of the TLR/NF-kB pathway, TNFAIP3 (A20) (Fig. 4c), which has been implicated by others in the prevention of asthma in rural populations 30 . While there was no increase in mRNA production with LPS exposure, there was an increase in A20 mRNA expression with HDE compared with control (p < 0.01). Interestingly, for HDE, this increase was inhibited by CO2 co-delivery at 1000 ppm (p < 0.05), and less inhibitory at 5000 ppm. mRNA expression of another protein shown to have effects on MAPK and NF-kB signaling, Hsp72 (Hsp1A) was also tested (Fig. 4d). CO2 at 5000 ppm co-exposure with saline was able to induce some Hsp72 increase in mRNA, but not a signi cant increase as compared to saline and room air. Co-exposed with room air, HDE induced a similar non-signi cant Hsp72 increase and LPS a signi cant increase (p < 0.01). However, Hsp72 mRNA was decreased when HDE or LPS stimuli were co-exposed with either CO2 concentration (p < 0.05 for LPS + CO2 treated groups).

Histology
Tissue sections (Fig. 5) overall showed little in the way of overt changes due to in ammation. This may be in part due to the limited time frame of exposure, and longer exposures are likely to produce more signi cant changes. No consistent overt differences were noted in mice for saline, LPS, or HDE treatments compared to their similarly treated but CO2 co-exposed duplicates.

Discussion
In high intensity animal production operations, there are buildups of both particulates in the air as well as a variety of gasses, mostly as a result of the animals housed there. Given the problems of respiratory effects in exposed workers 1,31 numerous studies have tried to recreate these conditions in animal models 2 to determine responsible agents for health effects. Given that CO2 elevation is common in many environments, not just in barns, we previously looked at the effects of workplace relevant (5000 ppm) CO2 on mice treated with barn dust extract to try to devise a more holistic model of barn exposure in ammation. The results from this study suggested that we look at both a more simple innate in ammatory molecule (LPS), as well as CO2 concentrations that are more re ective of a number of environments, making the results more meaningful for a variety of facilities, such as homes and o ces.
The rst questions was if breathing parameters showed any changes, as might be expected from an elevation in ambient CO2 levels. Others have reported potential stress after long repeated exposures to elevated CO2 in rats (700 ppm, 6hr/day over 15 days) 18 , whereas others have shown changes in minute volume associated with stressful procedures 32 . There were no signi cant changes to minute volume for any of the treated animals over the course of our studies, suggesting the treatments were reasonably tolerated. Similarly, breath frequency was unaffected across all treatments and CO2 levels. Tidal volume appeared to be the most sensitive of the tested parameters to treatments and CO2 exposure. Changes however appeared to be due to presence of CO2 alone as the secondary treatment had no effect. This is unsurprising as volume of air moved in and out of the lung should logically increase to counteract elevated CO2. Interestingly, within the HDE exposed animals the peak expiratory ow, peak inspiratory ow, and tidal volume measures increased with increasing ppm co-exposure to CO2, suggesting that when co-exposed, the increasing doses of carbon dioxide may be resulting in increases in airways responsiveness. While this would t with the barn dust extract and CO2 treatment being one of the most in ammatory treatments tested here, the changes were nearly immediate within the monitoring time (within 40 minutes, data not shown) which may suggest very early in ammatory mechanisms could be playing a role.
Looking at cellular levels in the lung extracellular space we see that similar to our previous study, increasing CO2 appeared to have no effect on saline treated animals. HDE by comparison showed a trend to increases in total cells and a trend to increased monocytes with CO2 co-treatment, even at levels as low as 1000 ppm CO2, and by 5000 ppm a signi cant increase in total BAL cells and neutrophils in the lung. This general increased in ammatory pro le is similar to our previous report 16 . A further similarity to that report was that in lung tissue histology (Fig. 5) there was no discernable changes seen with any of the treatments with or without added CO2. Given that this was a 6 hour exposure, we did not expect to see signi cant histological changes. We suspect a more chronic exposure may yield different results though, given other changes in cytokines and chemokine expression. The lack of signi cant physical changes to the lung also suggests that any changes to breathing parameters were likely not due to gross mechanical changes in the lungs over the course of the experiment.
Past work by our lab has shown that CO2 co-exposure did cause signi cant changes at the molecular level, increasing in ammatory responses to organic dusts. As LPS is a common constituent of agricultural dusts it was thought that LPS exposures and co-exposures may respond similarly to HDE. Our results showed a relatively common pattern across several cytokines with regards to LPS as compared to saline, showing signi cant increase in in ammatory cytokine production with LPS exposure. However, for the proin ammatory cytokines, the LPS with room air response tended to be greater than the HDE with room air response. These responses changed with co-exposure to CO2. For LPS, there were signi cant decreases in cytokines with both CO2 co-exposures. By contrast, when HDE was co-exposed with CO2, at 1000 ppm there was a decreased cytokine response compared to the room air HDE + CO2 coexposure, however, at 5000 ppm HDE + CO2 co-treatment the cytokines were greater as compared to the co-exposures to room air and 1000 ppm CO2. This results are similar to the BAL cellular responses, particularly with well-known monocyte and macrophage chemoattractors such as MCP-1, MIP-2, KC, TNFa, and IL-1B 33,34 . IL-1B should warrant particular attention given past work in ex-vivo systems with neutrophil microparticles that shows increased levels of IL-1b per microparticle with increasing CO2 up to 4000 ppm (Thom et al 2017). While we did not see this with unstimulated animals, we do see this increased IL-1B with HDE co-exposure. The similar results they show with human and mouse neutrophils in response to CO2 also suggest the mouse model does re ect what occurs in human cells. Their further proof of iNOS activation provides yet another avenue of exploration in our next studies.
Of the cytokines tested TNF-a, MCP-1, MIP-2 and KC are all are transcriptionally controlled by NF-kB and AP-1 (Nicholson 95, Redhu 11) [35][36][37][38] , suggesting a possible target for further testing as we try to deduce mechanisms of CO2 in ammation. Very little is currently known about a CO2 "sensor" or signaling pathway in cells. Several studies have reported effects on certain proteins such as ERK and JNK activation 39 . The most promising results for immunological effects have looked at changes to the IKKa and NF-kB alternate pathway signaling (P100 and RelB) in response to CO2 exposure 40,41 . These results implicate the NF-kB pathway as being involved. For this reason, we decided to examine expression of receptors and regulators of this pathway associated with organic dust exposure and LPS. TLR2 and TLR4 are two of the most heavily implicated receptors associated with HDE exposure 3,28,29,42 . TLR4 of course is also the primary receptor associated with LPS detection 43 . Interestingly we can see that TLR4 mRNA expression tended to be similar for all treatments. TLR2 on the other hand appears to exhibit a pattern of expression that follows a pattern similar to that seen with many of the proin ammatory cytokines we tested for, though other than for LPS + 1000 ppm CO2, the changes were not signi cant. As TLR2 can bind to LPS 44 , the reduction in TLR2 mRNA may suggest a mechanism for reduced LPS responses with higher CO2 but would not explain the increased indicators of in ammation seen with HDE. An examination at the mRNA and protein level of TLR2, TLR4, as well as MD-2 and CD14 43,45,46 , all of which are involved in the binding of LPS to TLR2 and TLR4 47 could assist in further delineating pathway involvements.
More recently an important upstream regulator of NF-kB signaling, TNFAIP3 (A20), was found to be vital in development of protection to asthma in rural children exposed to LPS 30 . While this may suggest that farm work may be protective against some respiratory problems, those working in concentrated animal feed operations receive high endotoxin exposure but show increased susceptibility to a variety of respiratory problems, including asthma 30 . Thus, we examined A20 mRNA expression in response to CO2.
While there was little expression of A20 when exposed to LPS alone, there was a signi cant increase in expression with exposure to organic dust, but this was inhibited by CO2 co-exposure, particularly at higher levels. While this result needs to be explored further it does suggest one way in which elevated CO2 may enhance HDE innate immune responses, by reducing expression of the inhibitor A20. Extrapolating to the previously mentioned study 30 such an A20 reduction may result in a non-protective exposure, if A20 is as critical to inhibiting future atopic responses as has been reported. This association of CO2 with worse asthma 48 and wheeze is shown in several daycare and school studies 10,11 . Finally, we looked at a very general heat shock protein, Hsp72 (HspA1). This particular protein has been shown to be produced in response to a broad array of stress stimuli 49 . Links between HspA1 and control of a more recently examined player in dust in ammation, HMGB1, made us interested in its induction. Interestingly, there were signi cant reductions in HspA1 for when LPS was co-exposed with CO2, even at levels as low as 1000 ppm CO2, and a trend in HDE treated animals in the same direction. As localization of HspA1 is a strong indicator of whether it is a pro or anti-in ammatory agent, longer exposures will need to be done in the future, looking at protein expression and localization. Our initial mRNA results suggest that addition of a CO2 co-exposure appears to reduce HspA1 mRNA when given in conjunction with an in ammatory stimulus.
While this work shows interesting results related to co-exposures to in ammatory agents, additional work is necessary. There is a need for a proteomic assessment of these in ammatory measures, to determine the breadth and scope of changes CO2 induces. Assessment of NF-kB signaling, including use of its alternate signaling pathway should be conducted. There is of course the need to assess chronic exposures, as well as situations of removing CO2 co-exposure at the start or end of an exposure period, to see if there is a time in which this increased CO2 exposure is critical to alteration of innate immunity.
Several of these parameters are part of ongoing work in our lab.
While this study was to look at exposures common in animal feed facilities, there is ample evidence that people are exposed in a number of environments to elevated CO2 that exceed 1000 ppm 15 . This work highlights that immunological exposures may not just be a factor of the introduced in ammatory agent, but also the environmental context in which it is given, and in which the subject lives and works. This may be particularly important where indications of respiratory symptoms in children such as wheeze, asthma 10,11 , and respiratory infection 50 could be exacerbated by increased CO2 exposure in their environment in addition to more commonly tested irritants. We show here that expression of mRNA of some innate immune receptors may be altered with just CO2 alone, but that response to in ammation can be signi cantly altered by these indoor CO2 levels at 1000 ppm or higher. Interestingly, our current study suggests that these responses will not necessarily be modi ed in the same manner across different in ammatory agents. This highlights the further need to deduce signaling pathways or commonalities in responses to certain immunological insults to better predict the impact of CO2. If response to an infection, or progression of certain chronic diseases (ex. COPD, asthma) can be impacted by elevated CO2, the possibility exists of ameliorating some effects of these same illnesses by improving ventilation.

Declarations
Ethics All experiments and procedures were approved by and conducted according to the protocols of the University of Saskatchewan Committee on Animals Care Assurance (AUP 20160104).

Consent for Publication
Not applicable

Availability of Data
The datasets generated during and/or analysed during the current study are available from the corresponding author on reasonable request.

Competing Interests
The authors declare they have no actual or potential competing nancial interests.

Funding
We would also like to thank the University of Saskatchewan College of Medicine Research Award (CoMRAD) for providing the nancial support for the presented study (awarded to "Effect of low dose Carbon Dioxide on barn dust and endotoxin induced Lung In ammation").
Author Contributions DS conceived the study and study design, performed treatments and most assays, collected and analyzed data. UP helped with data collection, analysis, and animal treatment. BT helped with data collection, sample processing and analysis of data and helped editing the manuscript. SK supervised the project and helped with data analysis, statistical measures, and in writing and editing the manuscript.