Preventing acute asthmatic symptoms by targeting a neuronal mechanism involving carotid body lysophosphatidic acid receptors

Asthma accounts for 380,000 deaths a year. Carotid body denervation has been shown to have a profound effect on airway hyper-responsiveness in animal models but a mechanistic explanation is lacking. Here we demonstrate, using a rat model of asthma (OVA-sensitized), that carotid body activation during airborne allergic provocation is caused by systemic release of lysophosphatidic acid (LPA). Carotid body activation by LPA involves TRPV1 and LPA-specific receptors, and induces parasympathetic (vagal) activity. We demonstrate that this activation is sufficient to cause acute bronchoconstriction. Moreover, we show that prophylactic administration of TRPV1 (AMG9810) and LPA (BrP-LPA) receptor antagonists prevents bradykinin-induced asthmatic bronchoconstriction and, if administered following allergen exposure, reduces the associated respiratory distress. Our discovery provides mechanistic insight into the critical roles of carotid body LPA receptors in allergen-induced respiratory distress and suggests alternate treatment options for asthma.

1. According to the authors a motivating factor of this study is to explain why carotid body resection is effective in human asthma. This seems somewhat disingenuous given the intense skepticism about any beneficial effect of carotid body resection on asthma. In fact, a position paper was published about this in JACI in August 1986. The authors point out that a review of the literature shows that when the patients were properly diagnosed and followed-up over time, there was no benefit of carotid body resection on asthma. Most of the studies are on unilateral resection, but the same skepticism is found in modern literature with the more dangerous bilateral resections. Therefore, I think it is misleading to presume the present data speaks to mechanisms underlying the efficacy of carotid body resection in human asthma.
If the paper would stick closer to the interesting basic science presented, and not so much on trying to prove asthma mechanisms, it would be an easier paper to follow. Bradykinin nebulization is referred to as "asthmatic provocation" in several places. The second sentence of the discussion referred to blocking LPA signaling as ameliorating acute asthmatic bronchoconstriction…" when I think "asthmatic bronchoconstriction" here refers to nebulized bradykinin-induced cholinergic reflex activity. The title refers to acute asthma attacks, when no such thing is studied, etc..

3.
A novel and very interesting aspect of this study is that it argues that the enhanced parasympathetic reflex response to nebulized bradykinin in allergically inflamed rat airways is abolished by acute denervation of the carotid bodies. This would mean that activity in the carotid sinus nerve is somehow required for the classic and well studied airway vagal C-fiber induced vagal parasympathetic bronchonconstictor reflex.
This would be very important, but also very peculiar. How do the authors imagine such a phenomenon occurring? The data presented do not make for a compelling case. To draw such a bold conclusion, more work is required. Is it possible that there is some artefactual effect of acutely ablating the carotid sinus nerves such that all vagal reflexes are blocked? Acutely severing inputs to the brainstem can have transient and profound effects. For example, bilateral severing of the vagus nerves in rodents can lead acutely to a complete loss of respiratory control and death. The hypothesis at hand is that the carotid bodies are needed for bradykinin-induced reflexes in allergically inflamed lungs. Might this be better addressed with a model whereby the carotid sinus nerves are cut in a survival surgery procedure. Then after the animals have had sufficient time to recover (~a week) one can determine if augmented vagal cholinergic airway reflexes are still blocked?
4. In the present study bradykinin, surprisingly, only evoked a reflex bronchoconstriction in allergically inflamed lungs? Does acute carotid sinus nerve ablation also prevent vagal airway reflexes in naïve animals. In other words, does acute severing of the carotid bodies abolish airway reflexes evoked by a stimulus that does not require airway inflammation to manifest itself? 5. The hypothesis here is that OVA leads to (a very modest) increase in plasma LPA and this is a necessary precondition for bradykinin applied to the lungs to evoke a bronchoconstrictor reflex (a reflex that somehow is dependent on actvity in the carotid bodies).
This conclusion is based only on the observation that BrP-LPA a potent LPA-receptor antagonist (with other actions such as phospholipase and phosphodiesterase inhibitory activity) abolished the ability of bradykinin to evoke a cholinergic reflex. Does BRP-LPA cross the BBB where it may interfere with neurotransmission? Perhaps phospholipase is needed for bradykinin to activate Cfiber terminals? Although the MS states bradykinin increases serum LPA in the OVA rats, the data in figure 3vi do not support this assertion. The baseline serum LPA is slightly elevated by OVA, but bradykinin does not seem to further enhance the LPA concentration. Is the LPA increased by the repetitive allergen challenge is somehow required for bradykinin to activate the airway C-fibers? Or is the argument that bradykinin effectively activated the airway C-fibers but can only lead to a cholinergic reflex if LPA in the serum facilitates carotid body activity. This is very confusing to meit would help if the authors stated their specific hypothesis and then presented the data in its favor. At a minimum one would wonder if more selective LPAr antagonists that are not enzyme inhibitors mimic the effect of BrP-LPA.
Since LPA receptors are expressed by the petrosal neurons, it would seem likely that they are also expressed nodose neurons. The vagal C-fibers terminate in the epithelium therefore these nerve endings are likely to come in contact with the largest concentration of the LPA-presuming, as stated, that OVA is stimulating LPA release from airway epithelium. Does nebulized LPA into the lungs evoke the reflex, as you argue that it will when infused to the CBs? If so is this airway reflex also blocked by lesioning the carotid sinus nerves. Can infused LPA into the lungs enhance the bradykinin evoked reflex or does it indeed need to reach the CBs?
We would like to thank the Reviewers for their time and consideration of our manuscript. We feel the reviewers have raised a number of important point which we have addressed below.
Reviewer #1 comments: 1. The concentrations of LPA used in different preparations and detected in rat plasma are in the micromolar range, which seems very high. Most prior studies reported that serum or lung LPA levels are in the nanomolar range (e.g. PMID 18583620, 24872406, 17359381, and PMC5521143). The use of ELISA to measure LPA is not in widespread practice, and many investigators prefer direct measurement using LC/MS. The authors need to justify their choice of ELISA and put their results in context with prior literature. It would be reassuring if they could confirm the high values of LPA observed using LC/MS, to complement the ELISA approach.
In light of this comment we have performed an extensive literature search of methods used to measure LPA and the concentration determined. Please see attached excel file.
LPA extended into the µM range in two of the four references cited by Reviewer 1 (see yellow highlighted in spreadsheet). In all we found 59 papers that measured LPA; 38 found LPA to be in the µM range. 30 used LC/MS,17 of which found LPA in the µM range --similar to that which we detected with ELISA --with the latest in 2017. 10 studies focused on lung disease; 3 of which found LPA extended into the µM range in blood, 1 which found LPA extended into the µM range in BALF, 2 of these studies used LC/MS. We attempted to the online portal, but the PDF conversion made readability difficult. To access an MS Excel version of this file please see: https://www.dropbox.com/s/m92rt4vvwj89p29/LPA%20concentration%20meta %20analyses.xlsx?dl=0 We choose to use ELISA because lipids are highly malleable and LC/MS involves multiple steps including a lipid extraction process which can be problematic. For example, alkenyl-acyl lipids in plasma can break down into LPA during the extraction process, which is still being optimized (Li et al. Int J Mol Sci 2014, 15, 10492-10507).
In recognizing Reviewer 1's concern however, we did additional validation experiments to confirm that asthma causes a physiological increase in plasma LPA sufficient to stimulate the carotid body: A) We developed a new bio-assay to test if the elevated LPA in plasma from asthmatic animals is sufficient to increase carotid body activity.
We took plasma from asthmatic rats 3hr after OVA challenge and perfused it through the en bloc carotid body preparation (1 ml of plasma in 100ml of perfusate; see revised manuscript for details, methods and Fig 4). Plasma from asthmatic rats increased carotid sinus nerve (CSN) activity by ~45% (plasma from naïve rats increased CSN activity by less than ~15%). Moreover, LPAr and TRPV1 antagonists abolished this increase. This experiment demonstrates the increase in plasma LPA translates to physiological effects on the carotid body. B) We further interrogated the accuracy of ELISA ( Supplementary Fig. 4 in revised manuscript). (a) We compared LPA calibration curves with and without plasma; plasma had no bearing on the outcome. (b) We compared LPA concentration in venous and arterial asthmatic plasma; there was no significant difference. (c) We compared EDTA and heparin processing of samples; there was no significant difference. (d) We evaluated the sensitivity of en bloc carotid body activity to additional species of LPA (16:0 and 18:2) that are more common in blood than the LPA we used previously (18:1); carotid body activity is sensitive to all three LPA species (Fig 1 e).
2. Most results seem to come from small numbers of animals per group (e.g. n=6). Please confirm that results have been confirmed in additional cohorts and replicates, which will help ensure reproducibility of results. This is a broad ranging study that has been further extended in addressing the comments of the reviewers. In all, the revised manuscript contains data from 234 animals (nearly double from the original manuscript), many of which are housed for 3+ weeks to develop the OVA model of asthma; these experiments are not trivial. We aim to minimize animal use, using repeated measure designs whenever possible, because of the time required for these experiments and our ethical and legal responsibilities. All our comparisons yielded p<0.01 except for borderline effects in evaluating dose responses.
With the effect size observed, power calculations demonstrate that in all of our experiments, the "n" was sufficient to obtain a power of 0.8 or greater.
We note that our judicious approach in selecting sample size is similar to and/or exceeds In Fig 4 (from original manuscript) where we examined the effects of LPA and TRPV1 antagonists on breathing in conscious animals, we originally used n=7. However, in analysing the data we realized we could perform a sub analysis to determine the longevity of the effect. In this sub-analysis the sample sizes dropped to 3 and 4 in each group. As this is a particularly important experiment, and heeding the reviewer's concern, we opted to increase the overall sample size to 12 allowing a sub analysis with groups of 6 and 6 (Please see Fig 6 in revised manuscript).
Please also note, in addressing Reviewer 2 comments we have also increased the n in experiments measuring LPA in plasma from asthmatic (7 to 13) and naïve rats (7 to  5. I recognize that the main focus of this paper is AHR and not inflammation per se. Are there any readouts of inflammation that can be shown in Figure 3? Done. We have taken two approaches to validate the inflammatory status of our model. First we calculated the Inflammation Index for the dataset described in Fig 3b from the original manuscript (see methods in revised manuscript). Second, we performed qPCR to measure gene expression of inflammatory mediators IL-4 and eotaxin on an additional cohort of asthmatic and naïve rats following bradykinin exposure. As expected, all measurements increased in the asthmatic group. These data are presented in Fig  This paper uses the ovalbumin sensitized and challenged Brown Norway rat as a model of allergic asthma. As with several other published studies, the data shows that these animals have an augmented vagal reflex response. In the present study the reflex response specifically to nebulized bradykinin is augmented in the allergically inflamed animals. Bradykinin is a strong activator of vagal C-fibers in the lungs of all laboratory animals. The increased response to bradykinin is therefore consistent with the findings of others that show vagal C-fiber responsiveness is nonselectively enhanced in this model e.g., Kuo and Lai have found that the C-fibers in the lungs are hyperresponsive to adenosine, capsaicin, 5-HT agonists in this same model (J. Appl. Physiol, 2008).
We thank the reviewer for this comment. As described in more detail below, it is important to distinguish between non-asthmatic and asthmatic airways. Both have robust responses to capsaicin, but only the asthmatic lung has a robust response to allergen and/or bradykinin (Sato, et al. 1996 General points 1. According to the authors a motivating factor of this study is to explain why carotid body resection is effective in human asthma. This seems somewhat disingenuous given the intense skepticism about any beneficial effect of carotid body resection on asthma. In fact, a position paper was published about this in JACI in August 1986. The authors point out that a review of the literature shows that when the patients were properly diagnosed and followedup over time, there was no benefit of carotid body resection on asthma. Most of the studies are on unilateral resection, but the same skepticism is found in modern literature with the more dangerous bilateral resections. Therefore, I think it is misleading to presume the present data speaks to mechanisms underlying the efficacy of carotid body resection in human asthma.
Done. We agree that carotid body denervation is highly controversial as a treatment of asthma and, even if it was efficacious in treating asthma, we absolutely do not advocate for its return to clinical practice because of the risk of death via asphyxiation. The Reviewer is correct that double-blinded clinical trials have demonstrated that unilateral denervation is without clinical efficacy. There are several reports that bilateral carotid body denervation reduces asthma severity, but these reports have not been confirmed by randomized control trials. However, animal work is unequivocal for an important role of the carotid body in bronchoconstriction and asthma ( If the paper would stick closer to the interesting basic science presented, and not so much on trying to prove asthma mechanisms, it would be an easier paper to follow. Bradykinin nebulization is referred to as "asthmatic provocation" in several places. The second sentence of the discussion referred to blocking LPA signaling as ameliorating acute asthmatic bronchoconstriction…" when I think "asthmatic bronchoconstriction" here refers to nebulized bradykinin-induced cholinergic reflex activity. The title refers to acute asthma attacks, when no such thing is studied, etc.. Done. We are intent on reaching a balance between scientific precision and the needs of the broad readership of Nature Communications. We have modified the body of the text to try and meet both goals and will consult the Editor regarding the wording of the title.
3. A novel and very interesting aspect of this study is that it argues that the enhanced parasympathetic reflex response to nebulized bradykinin in allergically inflamed rat airways is abolished by acute denervation of the carotid bodies. Correct.
This would mean that activity in the carotid sinus nerve is somehow required for the classic and well studied airway vagal C-fiber induced vagal parasympathetic bronchonconstictor reflex.
Incorrect. The reviewer is assuming that bradykinin is working exclusively through a C-fibre vagal-vagal reflex. In fact, bradykinin serves as a chemoattractant and stimulates mast cells that are in particular abundance in the asthmatic lung To demonstrate that the C-fibre induced vagal parasympathetic bronchonconstictor reflex is independent of the carotid body we performed a new set of experiments.
We challenged naïve carotid body denervated animals with capsaicin whilst measuring airway resistance using the Flexivent system. In one group of animals the vagus was intact and in another it was denervated. Capsaicin caused increased airway resistance in vagal intact but not vagotomised rats. Thus, the C-fibre mediated vagal-vagal reflex does not require the carotid bodies. This is in sharp contrast to the effects of nebulised bradykinin which is critically dependent on an intact carotid body (see Fig 5 d  The data presented do not make for a compelling case. To draw such a bold conclusion, more work is required. Is it possible that there is some artefactual effect of acutely ablating the carotid sinus nerves such that all vagal reflexes are blocked? Acutely severing inputs to the brainstem can have transient and profound effects. For example, bilateral severing of the vagus nerves in rodents can lead acutely to a complete loss of respiratory control and death. The hypothesis at hand is that the carotid bodies are needed for bradykinininduced reflexes in allergically inflamed lungs. Might this be better addressed with a model whereby the carotid sinus nerves are cut in a survival surgery procedure. Then after the animals have had sufficient time to recover (~a week) one can determine if augmented vagal cholinergic airway reflexes are still blocked?
Done. To address this comment, we performed additional experiments that involved asthmatic rats with chronically denervated carotid bodies. 4-5 days after denervation, we measured airway resistance in anesthetized rats using the Flexivent system. As in the acutely denervated preparation, bradykinin caused a pronounced increase in airway resistance in sham asthmatic rats but not in chronically denervated asthmatic rats (please see Fig 5c in  Does acute carotid sinus nerve ablation also prevent vagal airway reflexes in naïve animals. In other words, does acute severing of the carotid bodies abolish airway reflexes evoked by a stimulus that does not require airway inflammation to manifest itself?
As stated above, we challenged naïve carotid body denervated animals to capsaicin whilst measuring airway resistance using the Flexivent system. In one group of animals the vagus was intact and in another it was denervated. Capsaicin caused increase airway resistance in vagal intact but not vagotomised rats. Thus, the C-fibre mediated vagal-vagal reflex does not require the carotid bodies.
These data fit with Fig 2 h and i, in that LPA infusion in naïve animals produced carotid body-mediated bronchoconstriction and was lost with denervation.
5. The hypothesis here is that OVA leads to (a very modest) increase in plasma LPA and this is a necessary precondition for bradykinin applied to the lungs to evoke a bronchoconstrictor reflex (a reflex that somehow is dependent on activity in the carotid bodies).
With regards to the physiological relevance of the increase in plasma LPA, as reported above, we developed a new bio-assay to test if plasma containing asthma-elevated LPA was sufficient to increase carotid body activity. We took plasma from asthmatic rats 3hr after OVA challenge and perfused it through the en bloc carotid body preparation (1 ml of plasma in 100ml of perfusate; see revised manuscript for details). Plasma from asthmatic rats increased carotid sinus nerve (CSN) activity by ~45% (plasma from naïve rats increased CSN activity by less than ~15% ; Fig 4a,  . We confirmed this pathway using the dual perfused in situ and anesthetized preparations; carotid body activity causes an increase in vagal activity and increased airway resistance (Fig 2).
This conclusion is based only on the observation that BrP-LPA a potent LPAreceptor antagonist (with other actions such as phospholipase and phosphodiesterase inhibitory activity) abolished the ability of bradykinin to evoke a cholinergic reflex. Does BRP-LPA cross the BBB where it may interfere with neurotransmission? Perhaps phospholipase is needed for bradykinin to activate C-fiber terminals?
Done. As reported above, to address this important point, we performed additional experiments in the en bloc carotid body and the anesthetized in vivo preparations using the LPAR antagonist Ki16425 (LPA 1,3 and weak2 inhibitor). Ki16425 is not an phosphodiesterase inhibitor. Ki16425 and BrP-LPA had almost identical effects. These data are now added in Fig 1,  Although the MS states bradykinin increases serum LPA in the OVA rats, the data in figure 3vi do not support this assertion. The baseline serum LPA is slightly elevated by OVA, but bradykinin does not seem to further enhance the LPA concentration.
Done. As this was an important experiment and to comply with the spirit of Reviewer 1's request to consider group sizes we increased the number of animals in these experiments from a total of 14 to 28 animals. Analysis of the new data demonstrates a significant increase in LPA in response to bradykinin in OVA-treated animals (3.7±1.1; P<0.001). Figure 3e is updated accordingly along with a new supplementary Figure 4. Is the LPA increased by the repetitive allergen challenge is somehow required for bradykinin to activate the airway C-fibers? Or is the argument that bradykinin effectively activated the airway C-fibers but can only lead to a cholinergic reflex if LPA in the serum facilitates carotid body activity. This is very confusing to me-it would help if the authors stated their specific hypothesis and then presented the data in its favor.
Done. In the revised manuscript we have refined hypotheses and discussed both the carotid body-vagal reflex and the C-fibre vagal-vagal reflex with regards to their independence.
At a minimum one would wonder if more selective LPAr antagonists that are not enzyme inhibitors mimic the effect of BrP-LPA.
Done. Ki16425 had similar effects to Brp-LPA. Please see above for details.
Since LPA receptors are expressed by the petrosal neurons, it would seem likely that they are also expressed nodose neurons. The vagal C-fibers terminate in the epithelium therefore these nerve endings are likely to come in contact with the largest concentration of the LPA-presuming, as stated, that OVA is stimulating LPA release from airway epithelium.
Without disagreeing with these observations, we note that carotid body denervation abolished the ability of LPA to increase vagal activity in the in situ dual perfused preparation (Fig 2a-f). We also note that the effects of bradykinin on airway resistance in asthmatic (Fig 3d, 5b revised manuscript) and naïve anesthetized rats in response to intravenous LPA infusion (Fig 2h,  i) is abolished by carotid body denervation. Therefore, parsimony would suggest that the carotid bodies are the main site of action.
Does nebulized LPA into the lungs evoke the reflex, as you argue that it will when infused to the CBs? If so is this airway reflex also blocked by lesioning the carotid sinus nerves. Can infused LPA into the lungs enhance the bradykinin evoked reflex or does it indeed need to reach the CBs?
Done. To address this comment, we performed additional experiments in which we nebulized LPA into the lung of anesthetized asthmatic rats whilst measuring airway resistance using the Flexivent. Increase in airway resistance in response to LPA inhalation occurred after 30 min, congruent with a previous investigation (Hashimoto et al. 2001. Life Sci 70,[199][200][201][202][203][204][205]. Again, carotid body denervation minimized the increase in airway resistance. These data are consistent with a downstream effect of LPA on the carotid body. Please see the revised manuscript ( Supplementary Fig 5c in the revised  manuscript).
The authors have been quite responsive in the revised manuscript and rebuttal. New data are included: (1) demonstrating that plasma from asthmatic rats increased CSN activity in an LPARand TRPV1-dependent manner (FIg. 4), (2) validating the ELISA (Suppl. Fig. 4), and (3) demonstrating an effect of Ki16425 (i addition to Brp_LPA). These results strengthen their conclusions, and support the overall working model. I have a few comments and remaining concerns.
1. It is disappointing that the authors did not attempt to measure LPA concentrations or species using LC/MS. Not withstanding the concerns about potential artifacts introduced during extraction in the rebuttal, LC/MS remains the gold standard for lysolipid analyses and quantitation.
2. It would be helpful in Figure 6 cartoon to indicate when the antagonists were administered (similar to other figures).
3. Lines 271-278: The statements that "During allergen challenge, several species of LPA are released by lung epithelial cells into surrounding tissue...", and "....LPA is also released systemically, in arterial plasma" are not entirely correct and poorly worded. LPA is present in both circulation and also in lung fluids as sampled by BAL. The LPA species in these compartments are different, and although lung LPA levels increase after allergen challenge, it is not so clear that plasma levels also increase. In reference 30 (Park et al), the increase in serum LPA after allergen challenge was trivial. Whether LPA leaks between these different compartment (plasma and lung), or is generated de novo in different tissues, is not clear and an area of active research. This section should be re-worded and the references updated. Specific revisions should include and/or address the following.
-Reference 60 appears to be to a conference abstract and should be removed -Two human studies have shown that after allergen challenge, local concentrations of LPA increase in BAL fluids (Park et al, reference 30, and PMID 17359381 should also be included here) -In Park et al (reference 30) the increase in plasma LPA after allergen challenge was very small. The authors should acknowledge this, and point out that more research about changes in plasma LPA levels in human subjects after environmental exposures is needed in order to fully put their results in proper context. -Extracellular LPA can derive from phosphatidic acid (via the action of phospholipase) or from catalysis of LPC by the enzyme autotaxin (e.g. PMID 12354767) -LPA circulates bound to albumin and also autotaxin, and has a very short half-life due to the action of lipid phosphate phosphatases (LPP's) including LPP1 (PMID 19215222). Therefore, it is difficult to extrapolate from measurements of LPA to it's true bioactivity in vivo. -Factors responsible for increased LPA levels during inflammation (or in response to bradykinin in their model) are not known, and could reflect enhanced de novo generation or reduced breakdown.

No further comments
The authors have been quite responsive in the revised manuscript and rebuttal. New data are included: (1) demonstrating that plasma from asthmatic rats increased CSN activity in an LPAR-and TRPV1-dependent manner (Fig. 4), (2) validating the ELISA (Suppl. Fig. 4), and (3) demonstrating an effect of Ki16425 (in addition to Brp_LPA). These results strengthen their conclusions, and support the overall working model. I have a few comments and remaining concerns.
1. It is disappointing that the authors did not attempt to measure LPA concentrations or species using LC/MS. Not withstanding the concerns about potential artifacts introduced during extraction in the rebuttal, LC/MS remains the gold standard for lysolipid analyses and quantitation.
As suggested by Reviewer 1 and requested by the Editor we used LC/MS to measure LPA concentration in asthma plasma. We used our remaining frozen plasma samples extracted at the end of the experimental protocol in Fig 3 (the only point at which we could fully exsanguinate the animal). These new data confirm LPA in asthma plasma, ~20 min after challenge, is in the low micromolar range (1.5±0.2 µM).
The concentration measured with LC/MS is slightly lower than that measured with ELISA (8.4±1.7 µM). However, the relationship between LC/MS and ELISA measurements of the same samples was R= 0.8 (P=0.005). Confirming the quantitative capability of ELISA.
Our previous data using the en bloc carotid body preparation demonstrated 2.5 µM of LPA significantly increases carotid body activity; to ensure the physiological effect of LPA extends down to 1.5 µM we also add new data demonstrating 1.5 µM increases carotid body activity (supplementary Fig 5).
Please note that we used LC/MS as suggested by the reviewer notwithstanding the following important points: (a) That ELISA is used broadly in research and in FDA approved assays.  4b).
These experiments were acknowledged but not challenged by Reviewer 1. (e) In our revisions, we conducted a bioassay with our en bloc preparation to demonstrate that the plasma from challenged asthmatic rats increases carotid sinus nerve activity much more than plasma from non-asthmatic rats, and this difference was LPA receptor/TRPV1 dependent ( Figure 4). Thus, regardless of the difficulty in measuring the precise concentration of LPA in plasma because of its short half-life (~5 min; PMID: 23948545) and multiple mechanisms of degradation as described by Reviewer 1, the concentration of LPA in asthma plasma has biological effects on the carotid body. (f) Both our ELISA and LC/MS data may underestimate the peak LPA concentration following asthmatic challenge because blood was extracted ~20 min after challenge.
We have yet to determine the dynamics of LPA release and accumulation following challenge.
2. It would be helpful in Figure 6 cartoon to indicate when the antagonists were administered (similar to other figures). Done.
3. Lines 271-278: The statements that "During allergen challenge, several species of LPA are released by lung epithelial cells into surrounding tissue...", and "....LPA is also released systemically, in arterial plasma" are not entirely correct and poorly worded. LPA is present in both circulation and also in lung fluids as sampled by BAL. The LPA species in these compartments are different, and although lung LPA levels increase after allergen challenge, it is not so clear that plasma levels also increase. In reference 30 (Park et al), the increase in serum LPA after allergen challenge was trivial. Whether LPA leaks between these different compartment (plasma and lung), or is generated de novo in different tissues, is not clear and an area of active research. This section should be re-worded and the references updated. Specific revisions should include and/or address the following.
We thank the reviewer for these suggestions. Given our data we consider it highly likely that LPA stimulation of the carotid body originates from LPA released from the lung. However, acknowledging this reviewer's concerns and the fact that we have not explicitly determined that the lung is the source of increased LPA, we have subtly revised the manuscript accordingly. We have changed the wording of the title, modified the abstract and implemented the suggested changes below.
-Reference 60 appears to be to a conference abstract and should be removed Done.
-Two human studies have shown that after allergen challenge, local concentrations of LPA increase in BAL fluids (Park et al, reference 30, and PMID 17359381 should also be included here) Done.
-In Park et al (reference 30) the increase in plasma LPA after allergen challenge was very small. The authors should acknowledge this, and point out that more research about changes in plasma LPA levels in human subjects after environmental exposures is needed in order to fully put their results in proper context. Done.
-Extracellular LPA can derive from phosphatidic acid (via the action of phospholipase) or from catalysis of LPC by the enzyme autotaxin (e.g. PMID 12354767) Done.
-LPA circulates bound to albumin and also autotaxin, and has a very short half-life due to the action of lipid phosphate phosphatases (LPP's) including LPP1 (PMID 19215222). Therefore, it is difficult to extrapolate from measurements of LPA to it's true bioactivity in vivo. Done.
-Factors responsible for increased LPA levels during inflammation (or in response to bradykinin in their model) are not known, and could reflect enhanced de novo generation or reduced breakdown. Done.
Reviewer #2 (Remarks to the Author): No further comments