1. Neuroscience
Download icon

One bout of neonatal inflammation impairs adult respiratory motor plasticity in male and female rats

  1. Austin D Hocker
  2. Sarah A Beyeler
  3. Alyssa N Gardner
  4. Stephen M Johnson
  5. Jyoti J Watters
  6. Adrianne G Huxtable  Is a corresponding author
  1. University of Oregon, United States
  2. University of Wisconsin-Madison, United States
Research Article
  • Cited 0
  • Views 332
  • Annotations
Cite this article as: eLife 2019;8:e45399 doi: 10.7554/eLife.45399

Abstract

Neonatal inflammation is common and has lasting consequences for adult health. We investigated the lasting effects of a single bout of neonatal inflammation on adult respiratory control in the form of respiratory motor plasticity induced by acute intermittent hypoxia, which likely compensates and stabilizes breathing during injury or disease and has significant therapeutic potential. Lipopolysaccharide-induced inflammation at postnatal day four induced lasting impairments in two distinct pathways to adult respiratory plasticity in male and female rats. Despite a lack of adult pro-inflammatory gene expression or alterations in glial morphology, one mechanistic pathway to plasticity was restored by acute, adult anti-inflammatory treatment, suggesting ongoing inflammatory signaling after neonatal inflammation. An alternative pathway to plasticity was not restored by anti-inflammatory treatment, but was evoked by exogenous adenosine receptor agonism, suggesting upstream impairment, likely astrocytic-dependent. Thus, the respiratory control network is vulnerable to early-life inflammation, limiting respiratory compensation to adult disease or injury.

https://doi.org/10.7554/eLife.45399.001

eLife digest

Breathing is essential to life. At birth, the brain quickly adapts and learns to control breathing in different situations. This adaptability is called neuroplasticity. Most breathing-related adjustments in the brain are short-term, like breathing faster during exercise. The brain can also learn from prior experience to prepare for future situations. For example, intermittent exposure to low oxygen causes long-term changes in signals from the brain to muscles controlling breathing, which may help them prepare for future low oxygen situations. This is called long-term facilitation (LTF). This neuroplasticity may also help the brain to compensate or stabilize breathing during an illness or injury.

Illnesses shortly after birth can affect how the brain controls breathing and may contribute to respiratory diseases later in life. They may also have lasting effects on the ability to of the brain to learn and respond to stress, and may even contribute to psychiatric disorders or age-related cognitive decline.

Now, Hocker et al. show that inflammation shortly after birth has effects on breathing control that extend into adulthood. In the experiments, rats were injected four days after birth with either saline solution or a drug causing inflammation. When the rats grew into adults, their ability to make long-term breathing adjustments, or LTF, was assessed. In the rats exposed to early life inflammation two important pathways that enable LTF were eliminated. One pathway was restored when the rats received an anti-inflammatory treatment. Activating nerve cells reinstated the other pathway, suggesting these cells are not impaired.

The experiments suggest inflammation during early life impairs breathing control later on and may contribute to adult respiratory disease. Inflammation is common among infants in their first year, particularly among those born prematurely. This early-life inflammation may put them at risk of diseases associated with breathing control, like sleep apnoea, later in life. More studies are needed to understand the relationship between early life inflammation, respiratory control, and respiratory disease later in life.

https://doi.org/10.7554/eLife.45399.002

Introduction

At birth, neonates transition from a sterile maternal environment into an environment filled with pathogens, microbes, and toxins and must simultaneously begin robust, rhythmic breathing. Respiratory problems represent a significant clinical problem for neonatologists (Martin et al., 2012), especially in preterm infants where breathing is unstable (Poets and Southall, 1994; Poets et al., 1994) and infections are common (Stoll et al., 2002; Stoll et al., 2004). Further, inflammation appears to augment respiratory dysfunction in neonates, whereby inflammation depresses hypoxic responses (Olsson et al., 2003; Rourke et al., 2016) and induces recurrent apneas (Hofstetter et al., 2007). Despite the prevalence of early life inflammation, little is known about the long-lasting consequences of neonatal inflammation on adult neurorespiratory control.

We are beginning to understand the potential for long-term consequences of early life inflammation in other physiological systems. Neonatal inflammation blunts adult immune function (Bilbo et al., 2010; Mouihate et al., 2010; Spencer et al., 2011), increases adult stress reactivity (Shanks et al., 2000; Wang et al., 2013; Grace et al., 2014), impairs adult learning and hippocampal plasticity (Bilbo, 2005a; Bilbo et al., 2006), increases the risk of neuropsychiatric disorders (Rantakallio et al., 1997; Hornig et al., 1999), and worsens age-related cognitive decline (Bilbo, 2010). Yet, we know very little about the long-term effects of neonatal inflammation on adult neurorespiratory control.

Respiratory plasticity is an important feature of the neural control of breathing, providing adaptability and maintenance of breathing when the respiratory system is challenged (Fuller and Mitchell, 2017). Phrenic long-term facilitation (pLTF) is a frequently studied adult model of respiratory motor plasticity (Mitchell and Johnson, 2003) and is elicited by at least two distinct cellular signaling pathways: the Q-pathway and the S-pathway (reviewed in Dale-Nagle et al., 2010). The Q-pathway is evoked by moderate acute intermittent hypoxia (mAIH; 3 × 5 min hypoxic episodes, PaO235–45 mmHg) and is serotonin dependent, while the S-pathway is evoked by severe AIH (sAIH, PaO225–35 mmHg) and is adenosine dependent (Nichols et al., 2012). Interestingly, Q-pathway-evoked plasticity is undermined by even low levels of acute, adult, systemic inflammation and restored by the non-steroidal anti-inflammatory, ketoprofen (Vinit et al., 2011; Huxtable et al., 2013; Hocker and Huxtable, 2018), while S-pathway-evoked adult plasticity is inflammation resistant (Agosto-Marlin et al., 2017). Though we are beginning to understand more about the mechanisms of acute, adult inflammation on respiratory motor plasticity (Hocker et al., 2017; Hocker and Huxtable, 2018), we do not know how inflammation in early postnatal life impacts respiratory motor plasticity in the adult. Furthermore, few studies have investigated sex-differences in pLTF (Behan et al., 2002; Dougherty et al., 2017) and we know even less about sex-differences in respiratory control in response to inflammation. Additionally, males are more sensitive acutely to neonatal inflammation leading to higher male mortality in neonates (Bouman et al., 2005; Kentner et al., 2010; Rathod et al., 2017), but our understanding of other sex-differences after neonatal inflammation are unknown. Given the profound effects of neonatal inflammation on other physiological systems, we tested the hypothesis that neonatal inflammation undermines Q-pathway, but not S-pathway, respiratory motor plasticity in adult male and female rats.

Our results indicate that one neonatal inflammatory challenge completely abolishes adult, AIH-induced Q-pathway and S-pathway respiratory motor plasticity. Despite no lasting increases in adult, inflammatory gene expression, Q-pathway impairment is inflammation-dependent and is restored by acute adult anti-inflammatory treatment. Conversely, S-pathway impairment is inflammation-independent, but can be evoked by intermittent adenosine receptor agonism, suggesting phrenic motor neurons are not impaired. Since astrocytes are a primary source of adenosine during hypoxia (Takahashi et al., 2010; Angelova et al., 2015), they are likely impaired by neonatal inflammation and contributing to impairment of respiratory plasticity. These studies are the first steps toward understanding the lasting effects of neonatal inflammation on adult respiratory plasticity and suggest neonatal inflammation induces lasting-changes, increasing susceptibility to adult ventilatory control disorders.

Results

Neonatal inflammation acutely delays weight gain and increases male mortality

Male and female postnatal day 4 (P4) rats were injected with either LPS (Lipopolysaccharide; 1 mg/kg, i.p.) or saline. The dose of LPS was based on previous studies demonstrating CNS inflammatory gene expression in neonates (Rourke et al., 2016), as well as our unpublished data (N. Morrison, S. Johnson, J. Watters, A. Huxtable, unpublished observations). Within 24 hr of neonatal LPS injections, there was significantly greater mortality of male pups (8 of 67) than female pups (1 of 55, Fisher’s exact test, p = 0.04, Figure 1A). No mortality was evident in the saline treated males (n = 63) or females (n = 63). For the surviving pups, neonatal LPS males weighed significantly less at week 7 (no pairwise weight differences seen in females), but importantly, weights were not different in adults (Figure 1B).

Neonatal inflammation increases mortality in neonatal males and transiently delays weight gain in male and female rats.

After neonatal inflammation (P4, LPS 1 mg/kg, i.p.) male mortality (A) is increased within 24 hr (Fisher’s exact test, p = 0.006), but not in females (p = 0.466), relative to saline controls. Weekly male and female weights (B) after neonatal saline or LPS. (*p < 0.05, significant pairwise difference within sex).

https://doi.org/10.7554/eLife.45399.003

Adult, Q-pathway-evoked pLTF was undermined by neonatal inflammation, and restored by acute, adult anti-inflammatory treatment

Q-pathway-evoked pLTF is evident as the increase in integrated phrenic activity 60 min after mAIH (PaO235–45 mmHg) in adult, anesthetized rats (Bach and Mitchell, 1996). As expected, in adult males treated with neonatal saline, Q-pathway-evoked pLTF was evident after mAIH (55 ± 33.2% change from baseline, n = 7, p = 0.0006, Figure 2A and C). However, Q-pathway-evoked pLTF was absent in adult males treated with neonatal LPS (14 ± 49%, n = 12, p = 0.2247 Figure 2A and C). To control for the known effects of estrus cycle hormones on pLTF in females (Zabka et al., 2001; Behan et al., 2002; Dougherty et al., 2017), adult females were ovariectomized 7–8 days before electrophysiology studies. Similar to males, adult females treated with neonatal saline displayed Q-pathway-evoked pLTF (97 ± 63% change from baseline, n = 7, p < 0.0001, Figure 2B and C), while adult females challenged with neonatal LPS did not express pLTF (−15 ± 43%, n = 6, p = 0.4689, Figure 2B and C). Phrenic amplitude did not change from baseline in the time control group (8 ± 6% change, n = 5, p = 0.6482), regardless of sex or neonatal LPS exposure, and was significantly reduced compared to males or females treated with neonatal saline. Between groups, Q-pathway-evoked pLTF was significantly abolished in adults after neonatal LPS compared to adults after neonatal saline for both males (p = 0.0200) males and females (p < 0.0001). Thus, neonatal inflammation induces lasting impairment of adult, Q-pathway-evoked respiratory motor plasticity in both males and females.

Neonatal systemic inflammation undermines adult, Q-pathway-evoked pLTF in male and female rats.

Representative integrated phrenic neurograms from male (A) and female rats (B) after neonatal (P4) saline (top traces, black) or LPS (1 mg/kg, i.p.; bottom traces, grey). Q-pathway-evoked pLTF is evident in adults after neonatal saline as the progressive increase in phrenic nerve amplitude from baseline (dashed line) over 60 min following moderate acute intermittent hypoxia (mAIH, 3 × 5 min episodes, PaO235–45 mmHg). Group data (C) demonstrate Q-pathway-evoked pLTF 60 min after mAIH is abolished in adults by neonatal LPS in both males (circles) and females (triangles) and no change in phrenic amplitude in time controls (**p < 0.01, ***p < 0.001 from baseline, # p < 0.05, ### p < 0.001 between groups, ‡ p < 0.05 from adult males and females after neonatal saline).

https://doi.org/10.7554/eLife.45399.005

To test whether this lasting impairment of adult, respiratory motor plasticity was due to ongoing adult inflammation as a result of the neonatal inflammatory LPS challenge, we acutely treated adults with the non-steroidal anti-inflammatory, ketoprofen (12.5 mg/kg, i.p., 3 hr), a high dose previously shown to restore plasticity after acute, adult inflammation (Huxtable et al., 2013). Ketoprofen treatment restored Q-pathway-evoked pLTF in adult males treated with neonatal LPS (58 ± 18% change from baseline, n = 4, p = 0.0004, Figure 3A and C). Ketoprofen also restored Q-pathway-evoked pLTF in adult females treated with neonatal LPS (111 ± 44% from baseline, n = 5, p < 0.0001, Figure 3B and C). Adults treated with neonatal saline (male: 54 ± 17% from baseline, n = 4, 0.0008; female: 89 ± 40%, n = 5, p < 0.0001) were unaffected by adult ketoprofen treatment. Additionally, phrenic motor amplitude did not change in adult time controls treated with ketoprofen (13 ± 14% change from baseline, n = 4, p = 0.3436) and was significantly reduced compared to all other groups. Between groups, pLTF was not different between adult males (p = 0.7605) or females (p = 0. 2932) after neonatal saline or neonatal LPS, suggesting the impairment in Q-pathway-evoked pLTF is inflammation-dependent in both males and females.

Acute, adult anti-inflammatory (ketoprofen, Keto) restores Q-pathway-evoked pLTF after neonatal systemic inflammation in adult male and female rats.

Representative integrated phrenic neurograms for adult male (A) and female (B) rats after neonatal (P4) saline (top traces, black) or LPS (1 mg/kg, i.p.; bottom traces, grey) and acute, adult ketoprofen (12.5 mg/kg, i.p., 3 hr). Q-pathway-evoked pLTF is evident as the progressive increase in phrenic nerve amplitude from baseline (black dashed line) over 60 min following moderate acute intermittent hypoxia (mAIH, 3 × 5 min episodes, PaO235–45 mmHg). Group data (C) demonstrate adult ketoprofen restores Q-pathway-evoked pLTF 60 min after mAIH in adults after neonatal LPS in both males (circles) and females (triangles) and no change in phrenic amplitude in time controls (***p < 0.001 from baseline, ‡ p < 0.05 from all other groups).

https://doi.org/10.7554/eLife.45399.007

Neonatal inflammation did not induce chronic neuroinflammation in adult medulla or cervical spinal cords

Because the lasting impairment of Q-pathway-evoked pLTF was inflammation-dependent, we examined whether neonatal inflammation had lasting effects on adult neuroinflammation in regions involved in respiratory neural control and motor plasticity. Since plasticity was abolished in both males and females, data from both sexes were combined for analysis of inflammatory genes. In medullary and cervical spinal homogenates, neonatal LPS did not significantly alter mRNA for adult inflammatory genes (IL-6, IL-1β, TNF-α, or iNOS; Figure 4A and B). However, COX-2 gene expression was reduced in adult spinal cords after neonatal LPS (Figure 4B, p = 0.001), suggesting a decrease in COX-dependent inflammatory signaling. Thus, there was no evidence for lasting increases in neuroinflammatory gene expression in adults after a single exposure of neonatal inflammation in respiratory control regions.

Neonatal inflammation does not increase adult medullary or spinal inflammatory gene expression.

Homogenate samples isolated from adult medullas showed no significant increase in inflammatory mRNA after neonatal inflammation (A). Similarly, homogenate samples from isolated adult cervical spinal cords (B) were not increased by neonatal inflammation, but COX2 gene expression was significantly decreased in adults after neonatal inflammation (*p < 0.05).

https://doi.org/10.7554/eLife.45399.009

Adult S-pathway-evoked pLTF was undermined by neonatal inflammation, not restored by adult anti-inflammatory treatment, but revealed by intermittent adenosine 2A receptor agonism

S-pathway-evoked pLTF is evident as the increase in integrated phrenic activity 60 min after sAIH (PaO225–35 mmHg) in adult rats. As expected, in adult males after neonatal saline, S-pathway-evoked pLTF was evident after sAIH (61 ± 69% change from baseline, n = 5, p = 0.0001, Figure 5A and C). Contrary to our hypothesis, S-pathway-evoked pLTF was abolished in adult males after neonatal LPS (7 ± 18% change from baseline, n = 4, p = 0.6770, Figure 5A and C). In adult females after neonatal saline, S-pathway-evoked pLTF was evident (102 ± 47% change from baseline, n = 4, p < 0.0001, Figure 5B and C). Similar to adult males treated with neonatal LPS, S-pathway-evoked pLTF was abolished in adult females after neonatal LPS (0 ± 33%, n = 4, p = 0.9796, Figure 5B and C). Phrenic amplitude in the time control group was significantly less than males (p = 0.0147) or females (p < 0.0001) treated with neonatal saline. Between groups, S-pathway-evoked pLTF was significantly reduced after neonatal LPS in both adult males (p = 0.0180) and females (p < 0.0001) compared to adults after neonatal saline. Thus, neonatal inflammation induces lasting impairment of adult, S-pathway-evoked respiratory motor plasticity in both males and females.

Neonatal systemic inflammation undermines adult, S-pathway-evoked pLTF in male and female rats.

Representative integrated phrenic neurograms for adult male (A) and female (B) rats after neonatal (P4) saline (top traces, black) or LPS (1 mg/kg, i.p.; bottom traces, grey). S-pathway-evoked pLTF is evident as the progressive increase in phrenic nerve amplitude from baseline (black dashed line) over 60 min following severe acute intermittent hypoxia (sAIH, 3 × 5 min episodes, PaO225–35 mmHg) in adults after neonatal saline. Group data (C) demonstrate S-pathway-evoked pLTF 60 min after sAIH is abolished in adults by neonatal LPS in both males (circles) and females (triangles) and no change in phrenic amplitude in time controls (**p < 0.01, ***p < 0.001 from baseline ## p < 0.01, ### p < 0.001 between groups, ‡ p < 0.05 from male and female adults after neonatal saline).

https://doi.org/10.7554/eLife.45399.011

To test whether this lasting impairment of adult, S-pathway-evoked plasticity is due to ongoing inflammation in adults after neonatal LPS, we examined sAIH-induced plasticity after an acute, adult treatment with ketoprofen (12.5 mg/kg, i.p., 3 hr). Ketoprofen did not alter normal expression of S-pathway-evoked pLTF in adult males after neonatal saline (63 ± 22% change from baseline, n = 5, p = 0.0014, Figure 6A and C). However, contrary to the Q-pathway results, adult ketoprofen did not restore S-pathway-evoked pLTF in adult males after neonatal LPS (0 ± 65% change from baseline, n = 5, p = 0.9804, Figure 6A and C). Similarly, adult females treated with neonatal saline also exhibited normal S-pathway-evoked pLTF after adult ketoprofen (130 ± 22% change from baseline, n = 5, 0.0803, Figure 6B and C), and S-pathway-evoked pLTF was not restored by ketoprofen in adult females after neonatal LPS (25 ± 30% change from baseline, n = 6, p = 0.0803, Figure 6B and C). Adult males and females treated with neonatal LPS and adult ketoprofen were not different from time controls (males, p = 0.4964; females p = 0.5227). Between groups, S-pathway-evoked pLTF after acute ketoprofen was significantly reduced in adults after neonatal LPS comapred to adults after neonatal saline in both males (p = 0.0019) and females (p < 0.0001). Thus, neonatal inflammation induces a lasting impairment of adult, S-pathway-evoked respiratory motor plasticity, which is not due to ongoing adult, inflammatory signaling.

Adult, anti-inflammatory (ketoprofen, keto) does not restore S-pathway-evoked pLTF after neonatal systemic inflammation in adult male and female rats.

Representative integrated phrenic neurograms for adult male (A) and female (B) rats after neonatal (P4) saline (top traces, black) or LPS (1 mg/kg, i.p.; bottom traces, grey) and acute, adult ketoprofen (12.5 mg/kg, i.p., 3 hr). S-pathway-evoked pLTF is evident as the progressive increase in phrenic nerve amplitude from baseline (black dashed line) over 60 min following severe acute intermittent hypoxia (sAIH, 3 × 5 min episodes, PaO235–45 mmHg) in adults after neonatal saline. Group data (C) demonstrate acute, adult ketoprofen does not restore S-pathway-evoked pLTF 60 min after sAIH after neonatal LPS in adult males (circles) and females (triangles) and no change in phrenic amplitude in time controls (**p < 0.01, ***p < 0.001 from baseline, ## p < 0.01, ### p < 0.001 between groups, ‡ p < 0.05 from adult males and females after neonatal saline).

https://doi.org/10.7554/eLife.45399.013

S-pathway-evoked plasticity elicted by sAIH is adenosine dependent (Golder et al., 2008; Nichols et al., 2012) and can be evoked by intermittent CGS-21680, an adenosine 2A receptor agonist. To test if neonatal inflammation is impairing phrenic motor neurons and preventing pLTF, we examined phrenic output after intermittent CGS-21680 on the cervical spinal cord, around the phrenic motor pool. Intrathecal CGS-21680 (100 µM, 3 × 10 µL) evoked phrenic motor plasticity in adult males after neonatal saline (110 ± 17% change from baseline, n = 4, p < 0.001, Figure 7A and C) and females after neonatal saline (127 ± 47%, n = 4, p < 0.001, Figure 7B and C). After neonatal LPS, intrathecal CGS-21680 also elicited plasticity in adult males (85 ± 64%, n = 6, p < 0.001, Figure 7A and C) and adult females (147 ± 74%, n = 6, p < 0.001, Figure 7B and C), demonstrating phrenic motor neurons are not impaired after neonatal inflammation and are capable of S-pathway-evoked plasticity. The vehicle control group was not different from baseline (−3 ± 5% change, n = 4, p = 0.8891) and significantly reduced compared to all other groups. Between groups, pLTF was not different between adult males (p = 0.2841) or females (p = 0.4032) after neonatal saline or neonatal LPS. Thus, adult phrenic motor neurons are not impaired after neonatal inflammation and are capable of plasticity after neonatal inflammation. Therefore, the source of intermittent adenosine release is impaired during sAIH-induced pLTF after neonatal inflammation.

Intermittent adult, adenosine receptor agonism reveals plasticity after neonatal systemic inflammation in male and female rats.

Representative integrated phrenic neurograms for adult male (A) and female (B) rats after neonatal (P4) saline (top traces, black) or LPS (1 mg/kg, i.p.; bottom traces, grey). S-pathway-evoked phrenic motor plasticity is evident as the progressive increase in phrenic nerve amplitude from baseline (black dashed line) 90 min following intermittent CGS-21680 (100 µM, black arrows, 3 × 5 min apart) in adults after neonatal saline. Group data (C) demonstrate adult CGS-21680 reveals S-pathway-evoked plasticity after neonatal LPS in adult males (circles) and females (triangles) and no change in phrenic amplitude in vehicle controls (***p < 0.001 from baseline, ‡ p < 0.001 from adult males and females after neonatal saline).

https://doi.org/10.7554/eLife.45399.015

Adult microglia and astrocyte density were not changed by neonatal inflammation

While there was no evidence for elevated neuroinflammation based on the inflammatory genes evaluated here, the anti-inflammatory drug ketoprofen successfully restored Q-pathway-evoked plasticity. Additionally, our results indicate the impairment in S-pathway-evoked plasticity was likely due to a lasting change in adenosine signaling, possibly as a result of altered astrocytes. Thus, we hypothesized a lasting change in astrocytes and microglia in respiratory control regions, influencing neuronal function and impairing adult plasticity. We evaluated GFAP (astrocytes) and IBA1 (microglia) immunoreactivity in the adult preBötC, the site of respiratory rhythmogenesis (Smith et al., 1991), and in cervical spinal cords in the region of the phrenic motor nucleus, the presumptive site of pLTF (Baker-Herman et al., 2004; Devinney et al., 2015; Dale et al., 2017). Neonatal inflammation did not alter GFAP (p = 0.5969) or IBA1 (p = 0.6487) immunoreactivity in adult preBötC in either sex (Figure 8A,B and E), suggesting astrocyte and microglial density were not changed in adults after neonatal inflammation. Furthermore, there were no changes in GFAP (p = 0.7195) or IBA1 (p = 0.9254) immunoreactivity in adult cervical spinal cords (Figure 8C,D and F), suggesting no lasting changes in astrocyte and microglia density in the region of the phrenic motor nucleus. Additionally, no obvious differences in astrocyte or microglial morphology in adult phrenic motor nuclei or the preBötC were seen following neonatal LPS inflammation, suggesting other signaling mechanisms are responsible for impairing adult pLTF.

Neonatal inflammation does not alter GFAP or IBA1 immunofluorescence in adult preBötzinger Complex or ventral cervical spinal cords.

After neonatal LPS (1 mg/kg, i.p., (P4), representative confocal images (40x) from adult preBötC (A and B) and cervical spinal cords (C and D) displayed no qualitative differences in immunoreactivity for GFAP (green, astrocytes) or IBA1 (green, microglia) in males (left panels) or females (right panels). PreBötC neurons are labeled with antibodies for NK1R (red, A and B) and motor neurons are labeled with antibodies for ChAT (red, C and D). Neonatal inflammation did not significantly change mean fluorescent intensity of either GFAP or IBA1 in the preBötC (E) or cervical spinal cord (F), suggesting no lasting differences in astrocytes or microglia after neonatal inflammation. Scale bars: 50 µm.

https://doi.org/10.7554/eLife.45399.017

Acute hypoxic phrenic responses were greater in females, but were unaffected by neonatal inflammation

Neonatal inflammation did not significantly alter moderate acute hypoxic phrenic amplitude responses within adult males (neonatal saline = 114 ± 41% change from baseline; neonatal LPS = 93 ± 36%) or females (neonatal saline = 185 ± 53%; neonatal LPS = 148 ± 63%, Table 1). Hypoxic phrenic amplitude responses were also unaffected by the anti-inflammatory ketoprofen in adult males (neonatal saline +Keto = 118 ± 36%; neonatal LPS +Keto, 118 ± 44%) or females (neonatal saline +Keto, 165 ± 52%; neonatal LPS +Keto, 189 ± 82%, Table 1). However, adult females exhibited significantly greater acute phrenic amplitude responses to moderate hypoxia (main effect, p = 0.0004).

Table 1
Acute, adult hypoxic phrenic responses.
https://doi.org/10.7554/eLife.45399.019
MaleFemale †††
Neonatal salineNeonatal LPSNeonatal salineNeonatal LPS
Moderate hypoxia114 ± 4193 ± 36185 ± 53*148 ± 63
Keto + Moderate hypoxia118 ± 36118 ± 44165 ± 52189 ± 82*
MaleFemale †
Neonatal SalineNeonatal LPSNeonatal SalineNeonatal LPS
Severe hypoxia139 ± 37106 ± 10172 ± 125172 ± 26
Keto + Severe hypoxia151 ± 25174 ± 96194 ± 45235 ± 63
  1. Group data for adult, acute hypoxic phrenic responses to moderate (PaO235–45 mmHg) and severe (PaO225–35 mmHg) hypoxia demonstrate no differences after neonatal (P4) saline or LPS (1 mg/kg, i.p), or after adult ketoprofen (12.5 mg/kg, i.p, 3 hr) within each sex. Significant differences between sexes demonstrate larger responses in females after moderate or severe hypoxia († p<0.05, ††† p<0.001). *p<0.05 from male neonatal LPS. Moderate hypoxia: neonatal saline male (n = 7), neonatal LPS male (n = 10), neonatal saline female (n = 6), neonatal LPS female (n = 6). Keto + Moderate hypoxia: neonatal saline male (n = 4), neonatal LPS male (n = 4), neonatal saline female (n = 5), neonatal LPS female (n = 5). Severe hypoxia: neonatal saline male (n = 5), neonatal LPS male (n = 4), neonatal saline female (n = 4), neonatal LPS female (n = 4). Keto + Moderate hypoxia: neonatal saline male (n = 5), neonatal LPS male (n = 5), neonatal saline female (n = 5), neonatal LPS female (n = 6)

Phrenic amplitude in response to severe hypoxia was similarly unaltered by neonatal inflammation within adult males (neonatal saline = 139 ± 37% change from baseline; neonatal LPS = 106 ± 10%) or females (neonatal saline = 172 ± 125%; neonatal LPS = 172 ± 26%, Table 1). Acute ketoprofen pretreatment did not alter acute hypoxic phrenic amplitude responses in adult males (neonatal saline = 151 ± 25% change from baseline; neonatal LPS, 174 ± 96%) or females (neonatal saline = 194 ± 45%; neonatal LPS = 235 ± 63%, Table 1). Similarly, adult females exhibited a significantly greater acute amplitude response to severe hypoxia than males (main effect, p = 0.021).

Physiological parameters and frequency plasticity

All physiological parameters remained within experimental limits (Table 2). Neonatal saline or LPS caused no significant changes in adult temperature, PaCO2, PaO2, or pH at baseline. There were no between group differences in baseline MAP, suggesting no long-lasting cardiovascular changes after neonatal inflammation. No significant changes occurred over time in temperature, pH, or PaCO2 for any group. As expected, MAP and PaO2 were significantly decreased during hypoxic episodes in experimental groups (Huxtable et al., 2015; Hocker and Huxtable, 2018), but these changes were not evident in time control groups and were not different from baseline values at 60 min post-AIH.

Table 2
Physiological parameters during electrophysiology experiments.
https://doi.org/10.7554/eLife.45399.021
Temperature (°C)PaO2 (mmHg)PaCO2 (mmHg)pHMAP (mmHg)
BaselineMaleFemaleMaleFemaleMaleFemaleMaleFemaleMaleFemale
Neonatal Saline + mAIH37.4 ± 0.237.4 ± 0.2254 ± 19259 ± 5343.2 ± 5.648.9 ± 3.77.37 ± 0.067.36 ± 0.02124 ± 9121 ± 18
Neonatal LPS + mAIH37.6 ± 0.237.5 ± 0.2266 ± 30268 ± 2442.8 ± 4.7#45.0 ± 1.87.37 ± 0.047.37 ± 0.02127 ± 10123 ± 23
Neonatal Saline + Keto + mAIH37.3 ± 0.137.4 ± 0.3249 ± 21255 ± 2841.7 ± 4.948.6 ± 3.57.38 ± 0.037.33 ± 0.02132 ± 8121 ± 12
NeonatalLPS + Keto + mAIH37.4 ± 0.237.4 ± 0.3276 ± 40283 ± 3341.5 ± 3.247.9 ± 3.67.39 ± 0.037.34 ± 0.02129 ± 16117 ± 14
Neonatal Saline + sAIH37.6 ± 0.337.5 ± 0.2295 ± 18266 ± 943.3 ± 5.947.7 ± 3.17.37 ± 0.027.36 ± 0.00133 ± 6121 ± 19
Neonatal LPS + sAIH37.4 ± 0.137.4 ± 0.4297 ± 28264 ± 3545.3 ± 4.247.7 ± 4.07.36 ± 0.027.36 ± 0.02135 ± 20132 ± 14
Neonatal Saline + Keto + sAIH37.4 ± 0.337.5 ± 0.1256 ± 46268 ± 2941.6 ± 2.2e49.9 ± 3.57.39 ± 0.037.34 ± 0.02110 ± 12113 ± 29
NeonatalLPS + Keto + sAIH37.5 ± 0.237.4 ± 0.2243 ± 45245 ± 3942.3 ± 3.451.6 ± 6.87.37 ± 0.017.31 ± 0.04121 ± 3129 ± 13
NeonatalSaline + CGS-2168037.5 ± 0.237.3 ± 0.2268 ± 21237 ± 1445.4 ± 5.346.2 ± 37.36 ± 0.047.34 ± 0.03119 ± 24121 ± 10
Neonatal LPS + CGS-2168037.6 ± 0.337.4 ± 0.3247 ± 27247 ± 2146.2 ± 2.946.6 ± 2.77.37 ± 0.027.34 ± 0.02114 ± 12107 ± 15
Time Controls37.7 ± 0.1250 ± 4549.4 ± 4.07.36 ± 0.05110 ± 10
Time Controls + Keto37.5 ± 0.3250 ± 1744.0 ± 9.17.37 ± 0.06115 ± 39
CGS-21680 Vehicle Controls37.4 ± 0.3237 ± 4045.8 ± 0.87.37 ± 0.03110 ± 17
Hypoxia
Neonatal Saline + mAIH37.4 ± 0.237.3 ± 0.138 ± 2*,†, ‡40 ± 3*,†, ‡42.2 ± 5.1#48.8 ± 3.87.35 ± 0.077.36 ± 0.03#62 ± 23*,†, ‡65 ± 30*,†, ‡
Neonatal LPS + mAIH37.5 ± 0.437.4 ± 0.139 ± 2*,†, ‡39 ± 4*,†, ‡43.0 ± 4.8#45.4 ± 2.87.36 ± 0.04#7.36 ± 0.02#68 ± 19*,†, ‡80 ± 21
Neonatal Saline + Keto + mAIH37.5 ± 0.237.4 ± 0.338 ± 1*,†, ‡39 ± 3*,†, ‡41.9 ± 3.548.6 ± 3.77.37 ± 0.03#7.32 ± 0.0462 ± 1156 ± 6*,†, ‡
NeonatalLPS + Keto + mAIH37.5 ± 0.237.5 ± 0.340 ± 2*,†, ‡39 ± 4*,†, ‡40.9 ± 2.2#47.8 ± 4.97.36 ± 0.047.32 ± 0.0371 ± 1466 ± 19
Neonatal Saline + sAIH37.6 ± 0.237.4 ± 0.329 ± 5*,†, ‡29 ± 2*,†, ‡43.5 ± 5.947.7 ± 2.37.35 ± 0.037.32 ± 0.0658 ± 9*,†, ‡53 ± 11*,†, ‡
Neonatal LPS + sAIH37.4 ± 0.337.5 ± 0.330 ± 4*,†, ‡31 ± 5*,†, ‡46.3 ± 4.247.3 ± 5.97.34 ± 0.037.31 ± 0.0361 ± 2059 ± 20*,†, ‡
Neonatal Saline + Keto + sAIH37.4 ± 0.237.5 ± 0.230 ± 2*,†, ‡32 ± 3*,†, ‡42.3 ± 2.2#48.9 ± 3.77.36 ± 0.037.29 ± 0.0334 ± 8*,†, ‡,45 ± 29*,†, ‡
NeonatalLPS + Keto + sAIH37.3 ± 0.337.6 ± 0.231 ± 2*,†, ‡32 ± 1*,†, ‡42.2 ± 3.2#52.2 ± 5.87.34 ± 0.037.28 ± 0.0637 ± 11*,†, ‡,43 ± 21*,†, ‡
Time Controls37.6 ± 0.3226 ± 4048.7 ± 4.77.35 ± 0.04107 ± 13
Time Controls + Keto37.5 ± 0.2258 ± 1345.5 ± 9.37.37 ± 0.06#109 ± 44
60 min
Neonatal Saline + mAIH37.5 ± 0.437.3 ± 0.1234 ± 28259 ± 2243.4 ± 5.748.6 ± 3.77.38 ± 0.05#7.35 ± 0.02114 ± 9117 ± 27
Neonatal LPS + mAIH37.5 ± 0.337.4 ± 0.3253 ± 19268 ± 23*42.9 ± 4.4#45.2 ± 2.67.39 ± 0.04#7.37 ± 0.01116 ± 14121 ± 25
Neonatal Saline + Keto + mAIH37.3 ± 0.237.3 ± 0.3262 ± 14257 ± 3242.5 ± 4.948.7 ± 4.07.38 ± 0.017.33 ± 0.04121 ± 13115 ± 11
NeonatalLPS + Keto + mAIH37.6 ± 0.337.6 ± 0.3257 ± 18276 ± 40*41.5 ± 2.747.8 ± 3.67.36 ± 0.027.34 ± 0.06123 ± 11112 ± 18
Neonatal Saline + sAIH37.5 ± 0.337.4 ± 0.2262 ± 36258 ± 1943.7 ± 5.547.6 ± 2.97.37 ± 0.047.32 ± 0.03135 ± 9115 ± 24
Neonatal LPS + sAIH37.7 ± 0.237.4 ± 0.2282 ± 16*266 ± 1846.2 ± 4.847.7 ± 4.57.36 ± 0.027.35 ± 0.02127 ± 14128 ± 17
Neonatal Saline + Keto + sAIH37.7 ± 0.337.4 ± 0.3248 ± 42262 ± 942.3 ± 2.650.2 ± 4.27.37 ± 0.037.32 ± 0.03§105 ± 10109 ± 36
NeonatalLPS + Keto + sAIH37.5 ± 0.237.4 ± 0.3252 ± 24245 ± 2142.2 ± 351.2 ± 6.97.38 ± 0.027.31 ± 0.04117 ± 14125 ± 19
NeonatalSaline + CGS-2168037.3 ± 0.337.6 ± 0.1270 ± 46215 ± 4845.7 ± 4.947 ± 3.37.34 ± 0.067.35 ± 0.04112 ± 30126 ± 21
Neonatal LPS + CGS-2168037.4 ± 0.437.4 ± 0.4254 ± 26256 ± 2146.1 ± 3.246.5 ± 3.27.37 ± 0.017.33 ± 0.04107 ± 16101 ± 25
Time Controls37.5 ± 0.3220 ± 2548.4 ± 3.77.36 ± 0.04102 ± 22
Time Controls + Keto37.6 ± 0.2272 ± 2144.3 ± 8.77.37 ± 0.07111 ± 48
CGS-21680 Vehicle Controls37.6 ± 0.2271 ± 2745.4 ± 1.67.36 ± 0.03108 ± 7
  1. MAP, mean arterial pressure; PaO2, arterial oxygen pressure; PaCO2, arterial carbon dioxide pressure. Neonatal Saline +mAIH male (n = 7) female (n = 7); Neonatal LPS +mAIH male (n = 12) female (n = 6); Neonatal Saline +Keto + mAIH male (n = 4) female (n = 5); Neonatal LPS +Keto + mAIH male (n = 4) female (n = 5); Neonatal Saline +sAIH male (n = 5) female (n = 4); Neonatal LPS +sAIH male (n = 4) female (n = 4); Neonatal Saline +Keto + sAIH male (n = 5) female (n = 5); Neonatal LPS +Keto + sAIH male (n = 5) female (n = 6); Time Control (n = 5); Time Control + Keto (n = 4). Statistical comparisons: ANOVA-RM, Tukey’s post hoc: * different from Time control within time point, different from TC +Keto within time point,  different from baseline and 60 min, § different from baseline, # different from female neonatal LPS +Keto + sAIH within time point, different from female LPS within time point

Baseline phrenic burst frequency was not significantly different between groups and frequency plasticity, an increase in burst frequency 60 min after AIH (Baker-Herman and Mitchell, 2008), was not evident in any group. Phrenic burst frequency did not change after intrathecal CGS-21680.

Discussion

Although neonatal inflammation is common (Stoll et al., 2002; Stoll et al., 2004), little is known concerning how neonatal inflammation alters ventilatory control. Here, we investigated the long-term consequences of neonatal systemic inflammation on adult respiratory motor plasticity, a key feature of the neural control of breathing providing adaptability to respiratory system challenges (Mitchell and Johnson, 2003). We show for the first time that a single inflammatory challenge to neonates completely abolishes AIH-induced Q-pathway and S-pathway-evoked respiratory motor plasticity in adult males and females. Our results indicate a persistent change in adult inflammatory signaling contributes to this impairment since adult anti-inflammatory treatment restores Q-pathway-evoked, but not S-pathway-evoked, pLTF. Further, this is the first evidence of impairment of S-pathway-evoked motor plasticity, suggesting neonatal inflammation likely leads to a further vulnerable adult as this pathway was thought of as a ‘backup pathway’ after acute, adult inflammation (Agosto-Marlin et al., 2017). However, we demonstrate S-pathway plasticity can be revealed by intermittent, spinal adenosine receptor agonism, suggesting astrocyte dysfunction after neonatal inflammation since they are the likely source of adenosine during hypoxia (Takahashi et al., 2010; Angelova et al., 2015). These studies are the first steps toward understanding the lasting effects of neonatal inflammation on adult neurorespiratory control and suggest neonatal inflammation may increase susceptibility to adult ventilatory control disorders.

LPS-induced neonatal inflammation transiently upregulates cytokines in regions involved in respiratory control and plasticity (N. Morrison, S. Johnson, J. Watters, A. Huxtable, unpublished observations), consistent with inflammatory profiles in other CNS regions (Wang et al., 2006; Schwarz and Bilbo, 2011; Bilbo and Schwarz, 2012; Jafri et al., 2013). Neonatal inflammation also increases male mortality, consistent with clinical male mortality after neonatal inflammation (Person et al., 2014) and is relevant to the increased risk of sudden infant death syndrome for males (Kinney and Thach, 2009). However, similar to other studies (Bilbo et al., 2005b; Mouihate et al., 2010; Smith et al., 2014), we found no measurable adult changes in cytokines, or glial number or morphology despite a lasting inflammation-dependent impairment in Q-pathway-evoked pLTF. While we have previously demonstrated adult Q-pathway pLTF is sensitive to low levels of acute, systemic inflammation (Huxtable et al., 2011; Huxtable et al., 2013; Huxtable et al., 2015; Huxtable et al., 2018a; Vinit et al., 2011; Hocker and Huxtable, 2018), this is the first study to demonstrate neonatal inflammation induces lasting changes in inflammatory signaling to undermine adult Q-pathway-evoked pLTF. Additionally, S-pathway-evoked pLTF is not restored by even the high dose of ketoprofen used here, but is revealed by intermittent adenosine 2A receptor agonism on phrenic motor neurons (Seven et al., 2018). Thus, these results indicate phrenic motor neurons are not impaired after neonatal inflammation and the loss of S-pathway-evoked plasticity is likely due to impaired adenosine signaling during hypoxia. Furthermore, a likely source of adenosine during hypoxia is astrocytes (Takahashi et al., 2010; Angelova et al., 2015), suggesting neonatal inflammation induces lasting astrocyte-specific changes in the adult spinal cord to impair adult S-pathway-evoked pLTF. Further understanding the mechanisms impairing distinct forms of respiratory motor plasticity is required to develop plasticity as a therapeutic tool, such as for spinal cord injury and amyotrophic lateral sclerosis (Mitchell, 2008; Gonzalez-Rothi et al., 2015). Additionally, considering the cross-talk between Q- and S-pathways (Dale-Nagle et al., 2010; Fields and Mitchell, 2015; Perim et al., 2018), the response of the respiratory control system likely depends on the functional status of both Q- and S-pathways. Thus, future studies are needed to understand how inflammation modifies cross-talk between Q- and S-pathways and how respiratory motor plasticity can be exploited therapeutically (Gonzalez-Rothi et al., 2015).

The timing of neonatal inflammation is likely a significant factor in how neonatal inflammation impacts adult physiology. Low-levels of cytokines are important for neurodevelopment (Bilbo and Schwarz, 2009), and perturbing the balance of neonatal cytokines during development leads to lasting aberrant effects on neural circuits and developing cells (Reemst et al., 2016). Furthermore, while many components of the respiratory system begin developing in utero (Prakash et al., 2000; Pagliardini et al., 2003; Mantilla and Sieck, 2008; Johnson et al., 2018), the respiratory control system undergoes significant postnatal maturation. In these studies, we induced systemic inflammation with LPS at P4, similar to other studies showing long-term consequences of neonatal inflammation in other physiological systems (Shanks et al., 2000; Walker et al., 2006; Fan et al., 2008; Kohman et al., 2008; Bilbo, 2010), supporting the idea that important neural changes occur within the first week of life. Yet, it remains to be determined whether there is a precise critical period where neonatal inflammation impacts respiratory control circuits. However, our data on male mortality after neonatal LPS are consistent with other critical developmental windows, including a male-specific sensitive period to LPS (Rourke et al., 2016), disproportionate male mortality from neonatal inflammation (Person et al., 2014), the increased risk of sudden infant death syndrome for males (Kinney and Thach, 2009), and increased incidence of obstructive sleep apnea in adults after neonatal inflammation (McNamara and Sullivan, 2000). Thus, these data have important implications for understanding the sex-specific impairment early in life and into adulthood. Additionally, we and others (Spencer et al., 2006) observed a short delay in weight gain after neonatal inflammation, which normalized by weaning, suggesting no lasting effects on growth. Future studies are needed to refine our understanding of the critical periods during development when early-life inflammation induces long-lasting physiological changes to improve our understanding of adult disease and better understand important developmental processes.

While other reports have shown sex differences in neonatal programming of adult neuro-inflammatory responses (LaPrairie and Murphy, 2007; Rana et al., 2012), we observed no sex-differences in the effects of neonatal inflammation on adult plasticity. Importantly, this is the first evidence of inflammation abolishing pLTF in females and the first to report sAIH-induced respiratory motor plasticity in females. Females exhibited greater acute hypoxic phrenic amplitude responses relative to males, consistent with previous findings (Mortola and Saiki, 1996; Bavis et al., 2004), despite variability in reports of sex differences in hypoxic ventilatory responses (Behan and Kinkead, 2011). In contrast to our results following neonatal inflammation, neonatal stress alters adult hypoxic responses in a sex-dependent manner, whereby male responses are enhanced and female responses are blunted (Rousseau et al., 2017). Thus, the long-term effects on respiratory control may be dependent on the type of stressors in early life. Importantly, our experiments were performed in adult, ovariectomized females with exogenously restored estradiol levels to permit respiratory motor plasticity (Behan et al., 2002; Zabka et al., 2003; Dougherty et al., 2017). Therefore, as sex hormones are known to modulate respiratory control and hypoxic responses (Nelson et al., 2011; Behan and Kinkead, 2011), we cannot rule out a confounding role for exogenous estradiol supplementation after ovariectomy. Finally, after neonatal inflammation, we found no differences in adult hypoxic responses, suggesting no lasting change in carotid body responses due to neonatal inflammation. Accordingly, the deficit in adult respiratory motor plasticity after neonatal inflammation is likely a consequence of long-term changes in the spinal cord where pLTF occurs (Baker-Herman et al., 2004; Devinney et al., 2015; Dale et al., 2017).

While adult anti-inflammatory treatment restored Q-pathway-evoked pLTF, we did not observe increases in inflammatory gene expression in adult medullary or cervical spinal cord homogenates. Thus, while inflammatory signaling contributes to the impairment of adult plasticity, the source of this signaling change remains unclear and will be the topic of future studies. Similarly, others demonstrated no changes in baseline CNS inflammatory markers after neonatal inflammation, but observed priming of glial responses to adult stimuli (Bilbo et al., 2005b; Mouihate et al., 2010; Smith et al., 2014), suggesting lasting changes in glia have the potential to underlie impairments in adult respiratory plasticity. Contrary to other reports (Boissé et al., 2005; Kentner et al., 2010), we found spinal COX-2 gene expression was decreased in adulthood, suggesting a decrease in inflammatory signaling, which is unlikely to contribute to the lasting inflammation-dependent impairment in plasticity. Further, the acute inflammatory impairment of adult respiratory plasticity is COX-independent (Huxtable et al., 2018a), emphasizing a role for other inflammatory molecules mediating the lasting impairment in respiratory motor plasticity. Unmeasured inflammatory genes or post-transcriptional changes in inflammatory proteins may be responsible for undermining adult pLTF after neonatal inflammation. Conversely, other perinatal stimuli involving inflammatory signaling, such as maternal care and diet, do have lasting programming effects on adult inflammatory cytokine expression (Bilbo and Schwarz, 2009), but are more complex stimuli than the acute neonatal inflammation in our study. We also observed no change in microglial or astrocyte density and no obvious qualitative changes in morphology in adult medullas or spinal cords after neonatal inflammation. Thus, there are no obvious signs of inflammation in regions contributing to pLTF despite the restoration of Q-pathway-evoked pLTF with ketoprofen. Furthermore, the abolition of S-pathway-evoked pLTF is likely due to lasting changes in adenosine signaling from astrocytes, suggesting an astrocyte-specific change underlies this impairment. Thus, future studies are needed to identify inflammatory mechanisms undermining the Q-pathway and further details of the inflammation-independent mechanism responsible for undermining S-pathway-evoked motor plasticity.

The adult respiratory control network is vulnerable to early life stressors (Bavis et al., 2004; Genest et al., 2004; Fournier et al., 2011), which may undermine the ability to compensate during adult ventilatory control disorders. Our study is the first to demonstrate lasting consequences of neonatal inflammation on adult respiratory control. These deficits in respiratory control are independent of later life events, in contrast to other studies in which the physiological effects of early life inflammation are not revealed until after an adult stimulus (Bilbo et al., 2005b; Bilbo, 2010). We found a single episode of neonatal systemic inflammation induced lasting impairment of both Q- and S-pathway-evoked respiratory motor plasticity in adults. Our results suggest the adult impairment of Q-pathway plasticity is dependent on acute inflammatory signaling; however, we observed no lasting increase in adult inflammatory gene expression or the density of astrocytes and microglia. The pharmacological induction of S-pathway-evoked pLTF demonstrates phrenic motor neurons are capable of plasticity and suggest upstream impairment, such as the source of adenosine. While strong evidence supports astrocytes as the primary source of adenosine during hypoxia (Takahashi et al., 2010; Angelova et al., 2015), we cannot rule out other sources of adenosine. Identifying cell-type specific changes underlying lasting physiological impairments will be explored in future studies. Future studies will investigate the lasting effects of neonatal inflammation on isolated microglia and astrocytes to uncover potential mechanisms of adult impairments after neonatal inflammation.

Together, these results indicate two mechanistic pathways to spinal motor plasticity induced by AIH are undermined by neonatal inflammation in rats. Our experimental approach assessed phrenic nerve output in anesthetized rats and may not be generalizable to respiratory control in freely behaving animals or to other forms of motor plasticity. However, AIH induces long-term facilitation of ventilation in humans (Mateika and Komnenov, 2017) and strengthens corticospinal pathways to non-respiratory motor-neurons (Christiansen et al., 2018), suggesting our results likely have relevance to mechanisms of human spinal motor plasticity after AIH. While AIH-induced respiratory motor plasticity does not necessarily alter normal homeostatic control of ventilation, the general facilitation of spinal motor output has significant therapeutic potential for treating patients with respiratory and non-respiratory motor limitations (Trumbower et al., 2012; Trumbower et al., 2017; Nichols et al., 2013; Hayes et al., 2014)

In conclusion, this basic science study has major implications for the understanding the neonatal origins of adult ventilatory control disorders. These studies are the first evidence that one neonatal inflammatory exposure induces long-term impairments in adult respiratory control with potential relevance to many respiratory disorders. These findings are particularly relevant since inflammation is common in neonates (Person et al., 2014), especially those born prematurely who are at higher risk for adult disease (Luu et al., 2016). Improving our appreciation of how early life inflammation can influence adult respiratory control will have important consequences for understanding adult disease and susceptibility to respiratory disorders. Additionally, AIH-induced spinal motor plasticity is also a promising therapy to enhance motor recovery after spinal injury (Trumbower et al., 2012). However, not all patients respond to AIH (Hayes et al., 2014; Trumbower et al., 2017) and our findings suggest neonatal inflammatory exposure could contribute to these therapeutic limitations and understanding the mechanisms undermining plasticity will increase the therapeutic potential of AIH-induced spinal motor plasticity.

Materials and methods

Key resources table
Reagent type
(species) or
resource
DesignationSource or
reference
IdentifiersAdditional
information
Chemical compound, drugLPS (Lipopolysaccharides from e Coli (0111:B4))Sigma AldrichL4130dissolved in saline, 1 mg/ml
Chemical compound, drugKeto ((S) - (+) - Ketoprofen)Sigma Aldrich471909dissolved in 50% ethanol in saline,12.5 mg/ml
Chemical compound, drugCGS-21680Sigma AldrichC141dissolved in DMSO to50 mM for storage in aliquots. Dissolved to 100 uM in 10% DMSO and artificial CSF for injections.
Antibodyanti-GFAP (Rabbit polyclonal)Millipore(Millipore Cat# AB5804, RRID:AB_2109645)(1:1000)
Antibodyanti-NK1R (Guinea
pig polyclonal)
Millipore(Millipore Cat#
AB15810, RRID:AB_11213393)
(1:500)
Antibodyanti-IBA1 (Rabbit polyclonal)Wako(Wako Cat# 019–19741, RRID:AB_839504)(1:1000)
Antibodyanti-CHaT (Goat polyclonal)Millipore(Millipore Cat# AB144P, RRID:AB_2079751)(1:300)
Antibodydonkey-anti-rabbit 647 IgG secondaryLife Technologies(Molecular Probes Cat# A-31573, RRID:AB_2536183)(1:1000)
Antibodydonkey-anti-goat 555 IgG secondaryLife Technologies(Molecular Probes Cat# A-21432, RRID:AB_141788)(1:1000)
Antibodydonkey-anti-guinea pig 488 IgG secondaryJackson Immuno(Jackson ImmunoResearch Labs Cat# 706-545-148, RRID:AB_2340472)(1:1000)
Sequence-based reagentIL-1β forward primerIntegrated DNA TechnologiesCTG CAG ATG CAA TGG AAA GA
Sequence-based reagentIL-1β reverse primerIntegrated DNA TechnologiesTTG CTT CCA AGG CAG ACT TT
Sequence-based reagentIL-6 forward primerIntegrated DNA TechnologiesGTG GCT AAG GAC CAA GAC CA
Sequence-based reagentIL-6 reverse primerIntegrated DNA TechnologiesGGT TTG CCG AGT AGA CCT CA
Sequence-based reagentiNOS forward primerIntegrated DNA TechnologiesAGG GAG TGT TGT TCC AGG TG
Sequence-based reagentiNOS reverse primerIntegrated DNA TechnologiesTCT GCA GGA TGT CTT GAA CG
Sequence-based reagentTNFα forward primerIntegrated DNA TechnologiesTCC ATG GCC CAG ACC CTC ACA C
Sequence-based reagentTNFα reverse primerIntegrated DNA TechnologiesTCC GCT TGG TGG TTT GCT ACG
Sequence-based reagentCOX2 forward primerIntegrated DNA TechnologiesTGT TCC AAC CCA TGT CAA AA
Sequence-based reagentCOX2 reverse primerIntegrated DNA TechnologiesCGT AGA ATC CAG TCC GGG TA
Sequence-based reagent18 s forward primerIntegrated DNA TechnologiesCGG GTG CTC TTA GCT GAG TGT CCC
Sequence-based reagent18 s reverse primerIntegrated DNA TechnologiesCTC GGG CCT GCT TTG AAC AC

All experiments were approved by the Institutional Animal Care and Use Committees at the University of Oregon and the University of Wisconsin-Madison and conformed to the policies of the National Institute of Health Guide for the Care and Use of Laboratory Animals. Male and female Sprague Dawley rats (Envigo Colony 217 and 206) were housed under standard conditions (12:12 hr light/dark cycle) with food and water ad libitum.

Drugs and materials

LPS (0111:B4, Sigma Chemical) was dissolved and sonicated in sterile saline for neonatal intraperitoneal (i.p.) injections (1 mg/kg). S-(+) Ketoprofen (Keto, Sigma Chemical) was dissolved in ethanol (50%) and sterile saline for acute, adult injections (12.5 mg/ml/kg, i.p., 3 hr). 17-β estradiol was dissolved in sesame oil (Tex Lab Supply, Texas, USA) for acute injections (40 μg/mL/kg, i.p.,3 hr) in adult females after ovariectomy.

The adenosine 2A receptor agonist CGS-21680 was dissolved in fresh artificial cerebrospinal fluid (aCSF: 120 mM NaCl, 3 mM KCl, 2 mM CaCl2, 2 mM MgCl2, 23 mM NaHCO3, and 10 mM glucose) with DMSO (10%) for intrathecal injections.

Neonatal treatments

Timed pregnant rats (E14-17 upon arrival) were purchased in pairs from a commercial vendor (Envigo) and monitored daily. To control for between litter effects, litters were stratified such that each dam fostered similar numbers of male and female pups. On postnatal day 4 (P4), all of the stratified pups with one dam were injected with LPS (1 mg/kg, i.p.), while pups with the control dam were injected with sterile saline (i.p.). The dose of LPS was based on previous studies demonstrating CNS inflammatory gene expression in neonates (Rourke et al., 2016), as well as our unpublished data (N Morrison, S Johnson, J Watters, A Huxtable, unpublished observations) indicating CNS inflammation following LPS (1 mg/kg). Pups were weighed weekly and weaned at P21. Electrophysiology experiments were conducted once males reached 300 g. Females were ovariectomized at approximately 250 g, 7–8 days prior to electrophysiology experiments.

Ovariectomy

Ovariectomies were performed as previously described (Dougherty et al., 2017) to control for the known effects of estrus cycle hormones on pLTF (Zabka et al., 2001; Behan et al., 2002; Dougherty et al., 2017). Adult rats were anesthetized with isoflurane and maintained on a nose cone (2.5% in O2) during surgery. Depth of anesthesia was confirmed by the absence of toe-pinch responses. Bilateral dorsolateral incisions exposed ovarian fat pads. Ovaries were ligated and removed, muscle layers were approximated, and skin incisions were closed with a single dissolvable suture. A single dose of buprenorphine (0.05 g/kg, s.c.) was administered at the end of surgery for pain control and rats recovered in individual cages for 7-8 days before electrophysiology studies. Since pLTF exists in females only when estradiol is high (Dougherty et al., 2017), estradiol levels were restored by injection of 17-β estradiol (40 μg/mL/kg, i.p.) three hours before electrophysiology experiments.

Experimental groups

All experimental groups consisted of adult male and female rats after a single injection of either neonatal LPS or neonatal saline. To investigate the impact of neonatal systemic inflammation on adult Q-pathway-evoked respiratory motor plasticity, the following experimental groups were used: male neonatal saline + mAIH (n = 7), male neonatal LPS + mAIH (n = 12), female neonatal saline + mAIH (n = 7), female neonatal LPS + mAIH (n = 6).

To investigate if acute anti-inflammatory treatment restores Q-pathway-evoked respiratory motor plasticity after neonatal inflammation, adults were treated with ketoprofen (12.5 mg/kg, i.p.) three hours before electrophysiology experiments: male neonatal saline + Keto + mAIH (n = 4), male neonatal LPS + Keto + mAIH (n = 4), female neonatal saline + Keto + mAIH (n = 5), female neonatal LPS + Keto + mAIH (n = 5).

To investigate the impact of neonatal systemic inflammation on adult S-pathway-evoked respiratory motor plasticity, we used the following experimental groups: male neonatal saline + sAIH (n = 5), male neonatal LPS + sAIH (n = 4), female neonatal saline + sAIH (n = 4), female neonatal LPS + sAIH (n = 4).

To investigate if acute anti-inflammatory treatment restores S-pathway-evoked respiratory motor plasticity after neonatal inflammation, adults were treated with ketoprofen (12.5 mg/kg, i.p.) three hours before electrophysiology experiments: female neonatal saline + Keto + sAIH (n = 5), male neonatal LPS + Keto + sAIH (n = 5), female neonatal saline + Keto + sAIH (n = 5), female neonatal LPS + Keto + sAIH (n = 6).

To investigate if intermittent, intrathecal CGS-21680 reveals S-pathway-evoked respiratory motor plasticity, we used the following experimental groups: male neonatal saline + CGS-21680 (n = 4), male neonatal LPS + CGS-21680 (n = 6), female neonatal saline + CGS-21680 (n = 4), female neonatal LPS + CGS-21680 (n = 6).

To reduce use of additional animals, and because time control experiments were not statistically different between males or females, time control groups consisted of animals from each experimental condition. The time control group for studies investigating the Q-pathway (Figure 2) and S-pathway (Figure 5) consisted of adults after neonatal saline (male: n = 1; female n = 2), neonatal LPS (n = 1 male, 1 female). The time control + Keto group (Figures 3 and 6) consisted of adults after neonatal saline + Keto (n = 1 male, 1 female) and neonatal LPS + Keto (n = 1 male, 1 female). Vehicle controls for intrathecal CGS-21680 experiments (Figure 7) consisted of adults after neonatal saline (n = 1 male, 1 female) and neonatal LPS (n = 1 male, 1 female).

Electrophysiological studies

Electrophysiological studies have been described in detail previously (Bach and Mitchell, 1996; Baker-Herman and Mitchell, 2002; Huxtable et al., 2013). Rats were anesthetized with isoflurane, tracheotomized, ventilated (Rat Ventilator, VetEquip), and vagotomized bilaterally. A venous catheter was placed for drug delivery and fluid replacement, and a femoral arterial catheter was used to monitor blood pressure and for arterial blood sampling. Arterial blood samples were analyzed (PaO2, PaCO2, pH, base excess; Siemens RAPIDLAB 248) during baseline, during the first hypoxic response, and 15, 30 and 60 min post-AIH. Temperature was measured with a rectal temperature probe (Kent Scientific Corporation) and maintained between 37°C and 38°C with a custom heated table. Using a dorsal approach, hypoglossal and phrenic nerves were cut distally, and de-sheathed. Rats were converted to urethane anesthesia (1.8 g/kg i.v.; Sigma-Aldrich), allowed to stabilize for one hour, and paralyzed with pancuronium dibromide (1 mg; Selleck Chemicals).

In rats receiving intrathecal injections, a laminectomy was performed at cervical vertebrae 2 (C2) and a primed, silicone catheter was inserted two millimeters through a small incision in the dura. The catheter tip extended toward the rostral margin of C4 (Baker-Herman and Mitchell, 2002). CGS-21680 (100 μM) or vehicle (10% DMSO in aCSF) was injected around the phrenic motor pool in three boluses (10 µL) separated by 5 min.

Nerves were bathed in mineral oil and placed on bipolar silver wire electrodes. Raw nerve recordings were amplified (10 k), filtered (0.1–5 kHz), integrated (50 ms time constant), and recorded (10 kHz sampling rate) for offline analysis (PowerLab and LabChart 8.0, AD Instruments). Apneic and recruitment CO2 thresholds were determined by changing inspired CO2 with continuous end-tidal CO2 monitoring (Kent Scientific Corporation). End tidal CO2 was set 2 mmHg above the recruitment threshold, whereby arterial blood samples were used to establish baseline PaCO2, which was maintained within 1.5 mmHg of the baseline value throughout. Blood volume and base excess were maintained (±3 MEq/L) by continuous infusion (1–3 mL/h, i.v.) of hetastarch (0.3%) and sodium bicarbonate (0.99%) in lactated ringers. Experiments were excluded if mean arterial pressure deviated more than 20 mmHg from baseline.

All rats (excluding time control rats) received three, 5 min bouts of either mAIH (~10.5% O2, PaO235–45 mmHg) or sAIH (~7% O2, PaO225–35 mmHg). The average amplitude and frequency of 30 consecutive integrated phrenic bursts were taken during baseline, the first acute hypoxic response, and 15, 30, and 60 min after AIH and made relative to baseline amplitude. Phrenic nerve activity data for each experimental group were compared using two-way, repeated measures ANOVA with Fisher LSD post hoc tests. Sample sizes were selected based on similar, previous studies and the variance of pLTF in our experience (Huxtable et al., 2018a; Huxtable et al., 2018b; Hocker and Huxtable, 2018). Physiological variables were compared using two-way, repeated measures ANOVA with Tukey’s post hoc test. Mean arterial pressure is reported for baseline, the end of the third hypoxic exposure, and 60 min after AIH. Acute hypoxic responses were compared using an ANOVA with Fisher LSD post hoc test. Values are means ± SD.

RNA isolation, cDNA synthesis and quantitative PCR experiments

Neonatal rats (P4) were injected with either vehicle (saline) or LPS (1 mg/kg, i.p.) and allowed to mature to ~12 weeks. Adult male and female rats were anesthetized with isoflurane and perfused with PBS (transcardiac). Medulla and cervical spinal cords (C3-C7) were dissected and flash frozen until they were homogenized in Tri-Reagent (Sigma, St. Louis, MO, USA). Glycoblue reagent (Invitrogen, Carlsbad, CA, USA) was used to isolate total RNA, according to the manufacturer’s protocol. cDNA was reverse transcribed from 1 µg of total RNA using MMLV reverse transcriptase together with a cocktail of oligo dT and random primers (Promega, Madison, WI, USA), as previously described (Crain and Watters, 2015), and analyzed using qPCR with PowerSYBR green PCR master mix on an ABI 7500 Fast system. Inflammatory gene expression was analyzed in medulla and spinal cord homogenates using the following primers:

  • IL-6: 5’-GTG GCT AAG GAC CAA GAC CA and 5’-GGT TTG CCG AGT AGA CCT CA;

  • IL-1β: 5’-CTG CAG ATG CAA TGG AAA GA and 5’-TTG CTT CCA AGG CAG ACT TT;

  • COX-2: 5’-TGT TCC AAC CCA TGT CAA AA and 5’-CGT AGA ATC CAG TCC GGG TA;

  • TNF-α: 5’-TCC ATG GCC CAG ACC CTC ACA C and 5’-TCC GCT TGG TGG TTT GCT ACG;

  • iNOS: 5’-AGG GAG TGT TGT TCC AGG TG and 5’-TCT GCA GGA TGT CTT GAA CG;

  • 18 s: 5’-CGG GTG CTC TTA GCT GAG TGT CCC G and 5’-CTC GGG CCT GCT TTG AAC AC.

Wherever possible, primers were designed to span introns (Primer three software) and were purchased from Integrated DNA Technologies (Coralville, IA, USA). Primer efficiency was assessed by use of standard curves, as previously reported (Crain and Watters, 2015). Expression of inflammatory genes was made relative to 18 s ribosomal RNA calculated using the 2-ΔΔCT method (Livak and Schmittgen, 2001). Gene transcripts were considered undetectable, and not included in statistical analyses if their CT values fell outside of the linear range of the standard curve for that primer set, which in most cases was ≥34 cycles.

Immunohistochemistry methods

Upon completion of electrophysiology experiments, rats were perfused (transcardiac) with cold phosphate buffered saline (PBS, pH 7.4), followed by 4% paraformaldehyde (pH 7.4). All brains were removed and immersed in paraformaldehyde until sectioning (Leica VT 1200S vibratome). For immunohistochemistry, transverse medullary and coronal cervical spinal cord sections (40 µm) were washed (PBS) and blocked (PBS, 0.3% Triton, 1% BSA, 2 hr, room temperature) to prevent non-specific antibody binding. For medullary sections, two combinations of primary antibodies were used (PBS, 0.3% Triton, 0.01% BSA, room temperature, 24 hr): (1) rabbit anti-GFAP (1:1000, Millipore AB5804) to label astrocytes and guinea pig anti-NK1R (1:500, Millipore AB15810) to label preBötzinger Complex (preBötC) neurons (Gray et al., 1999), and (2) rabbit anti-IBA1 (1:1000, Wako 019–19741) to label microglia and guinea pig anti-NK1R (1:500, Millipore AB15810) to label preBötC neurons. For the spinal cord, two different combinations of primary antibodies were used (PBS, 0.3% Triton, 0.01% BSA, room temperature, 24 hr): (1) rabbit anti-GFAP (1:1000, Millipore AB5804) to label astrocytes and goat anti-ChAT (1:300, Millipore AB144p) to label motor neurons, (2) rabbit anti-IBA1 (1:1000, Wako 019–19741) to label microglia and goat anti-ChAT (1:300, Millipore AB144p) to label motor neurons. After primary antibody incubation, sections were rinsed (PBS) and incubated with secondary antibodies (PBS, 0.3% triton, 0.01% BSA, room temperature, 3 hr): donkey-anti-rabbit 647 IgG (1:1000, Life Technologies A31573) to label GFAP and IBA1 primary antibodies, donkey-anti-goat 555 IgG (1:1000, Life Technologies A21432) to label ChAT primary antibody and donkey-anti-guinea pig 488 IgG (1:1000, Alexa Fluor 706-545-148) to label NK1R primary antibody. Sections were washed and mounted onto charged microscope slides, air dried and covered with prolong gold (Life technologies, P36930) to preserve the fluorescence. A glass cover slip was placed over the samples and sealed with clear nail polish. Slides were stored in the dark at 4°C until imaged. All immunohistochemistry experiments contained adult male and female tissues after neonatal saline (medulla: n = 5 males, seven females; spinal cord: n = 5 males, six females) or neonatal LPS (medulla: n = 6 males, four females; spinal cord: n = 6 males, three females).

Image analysis methods

All immunofluorescent images (1024 × 1024 pixels, 40x magnification) were acquired using a Leica Microsystems CMS GmbH confocal microscope using the LAS X acquisition and viewing software (0.5 µm z-stack step increments). All images were taken using identical laser and gain settings and identically adjusted for contrast/brightness using ImageJ open source software to allow for comparisons across all groups. To quantify the density of microglia and astrocytes, maximum intensity projections for 20 µm of z-stacks from the medulla and cervical spinal cords were analyzed. Mean fluorescent intensity for each image within a single batch was made relative to the average fluorescent intensities of adults after neonatal saline samples within each sex (Paizs et al., 2009). Data are presented as percent change from adults after neonatal saline within each sex.

Statistical analysis

GraphPad Prism 7.0 software was used for statistical analyses. Differences in mortality between treatments and between sexes was evaluated with Fisher’s exact test. Phrenic nerve activity data for each experimental group were compared using two-way, repeated measures ANOVA with Fisher LSD post hoc tests. Physiological variables were compared using two-way, repeated measures ANOVA with Tukey’s post hoc test. Mean arterial pressure is reported from baseline, the end of the third hypoxic exposure, and 60 min after AIH. Acute hypoxic phrenic responses were compared using an ANOVA with Fisher LSD post hoc test. Microglial and astrocytic density comparisons were made between groups using a one-way ANOVA with multiple-comparisons post hoc tests. For all tests, p < 0.05 was considered significant and all data are expressed as mean ± SD.

References

  1. 1
  2. 2
  3. 3
  4. 4
  5. 5
  6. 6
  7. 7
  8. 8
  9. 9
  10. 10
  11. 11
  12. 12
  13. 13
  14. 14
  15. 15
  16. 16
  17. 17
  18. 18
  19. 19
  20. 20
  21. 21
  22. 22
  23. 23
  24. 24
  25. 25
  26. 26
  27. 27
  28. 28
  29. 29
  30. 30
  31. 31
  32. 32
  33. 33
  34. 34
  35. 35
  36. 36
  37. 37
  38. 38
  39. 39
  40. 40
  41. 41
  42. 42
  43. 43
  44. 44
  45. 45
  46. 46
  47. 47
  48. 48
  49. 49
  50. 50
  51. 51
  52. 52
  53. 53
  54. 54
  55. 55
  56. 56
    Genetic Basis for Respiratory Control Disorders
    1. GS Mitchell
    (2008)
    291–311, Respiratory plasticity following intermittent hypoxia: A guide for novel therapeutic approaches to ventilatory control disorders?, Genetic Basis for Respiratory Control Disorders, Springer, 10.1007/978-0-387-70765-5_17.
  57. 57
  58. 58
  59. 59
  60. 60
  61. 61
  62. 62
  63. 63
  64. 64
  65. 65
  66. 66
  67. 67
  68. 68
  69. 69
  70. 70
  71. 71
  72. 72
  73. 73
  74. 74
  75. 75
  76. 76
  77. 77
  78. 78
  79. 79
  80. 80
  81. 81
  82. 82
  83. 83
  84. 84
  85. 85
  86. 86
  87. 87
  88. 88
  89. 89
  90. 90
  91. 91
  92. 92
  93. 93
  94. 94

Decision letter

  1. Jan-Marino Ramirez
    Reviewing Editor; Seattle Children's Research Institute and University of Washington, United States
  2. Ronald L Calabrese
    Senior Editor; Emory University, United States
  3. Jan-Marino Ramirez
    Reviewer; Seattle Children's Research Institute and University of Washington, United States

In the interests of transparency, eLife includes the editorial decision letter and accompanying author responses. A lightly edited version of the letter sent to the authors after peer review is shown, indicating the most substantive concerns; minor comments are not usually included.

[Editors’ note: a previous version of this study was rejected after peer review, but the authors submitted for reconsideration. The first decision letter after peer review is shown below.]

Thank you for submitting your work entitled "A single bout of neonatal inflammation impairs adult respiratory motor plasticity in male and female rats" for consideration by eLife. Your article has been reviewed by three peer reviewers, including Jan-Marino Ramirez as the Reviewing Editor and Reviewer #1, and the evaluation has been overseen by a Senior Editor.

Our decision has been reached after consultation between the reviewers. Based on these discussions and the individual reviews below, we regret to inform you that your work will not be considered further for publication in eLife.

The reviewers felt that additional experiments will be necessary to increase the significance of your findings. The reviewers had several specific recommendations and suggestions.

1) It would be important to demonstrate that there is indeed significant inflammation in areas relevant for LTF – e.g. phrenic nucleus – 48hrs after LPS injection and 1 or two weeks later.

2) The reviewers also recommended to add some experiments that show that LTF translates into an altered hypoxic ventilatory response in freely moving unanesthetized animal.

3) Another issue that was raised relates to the apparent lack of a glial involvement. It would be more convincing to the readers, if the authors would add a deeper glial characterization e.g. trough flow cytometry, an approach which you seem to be familiar with (Nikodemova et al., 2012; 2014; 2015; 2016), and/or an additional test to explore whether glial responsivity to inflammatory challenges has been altered (Li et al., 2010; Cai et al., 2013; Yin et al., 2013; Lunardelli et al., 2015). It is possible that the glia seem unchanged under "resting" conditions but become more prone (primed) towards activation, something you are also very familiar with (Huxtable et al., 2013; Johnson et al., 2018).

Reviewer #1:

This is a solid study by Hocker et al., investigating the consequences of neonatal LPS injection on respiratory plasticity. The authors report several interesting findings.

One finding is the increased neonatal mortality observed in male versus female pups. This is a very important finding in the context of SIDS, which affects males significantly more than female. The ratio of mortality seen in this study is similar to that seen in human SIDS. The only minor concern is that this interesting observation is not further explored. Hence it seems a bit of an "add-on" rather than an integral part of the story reported in this manuscript.

The remainder of the manuscript focuses on respiratory plasticity. It follows in the footsteps of the Gordon Mitchell laboratory and carefully dissects the so called "Longterm facilitation" into the different cellular pathways and the author characterize the underlying mechanisms of the LTF as well as the changes in female and males exposed to neonatal LPS.

The single bolus of LPS has long-term differential consequences on this form of plasticity and the underlying mechanisms. The authors perform e.g. a careful analysis of the cytokines involved – and the involvement is rather diverse. This is perhaps not surprising.

The one finding that seems unexpected is that the authors cannot identify the actual mechanism underlying the long-term impairment of the plasticity. They did not discover any changes in astrocytes and microglia, neither abundance of these cells nor in their morphology. This "negative finding" leaves the reader somewhat "unsatisfied" as it makes the study appear somewhat descriptive and "unresolved". However, this may not be the case – the authors have done a great job in unraveling the subcellular mechanisms underlying the plasticity and the changes associated with the one-time inflammatory event. Perhaps the lack of glial alterations is the most important finding, i.e. that the plasticity is disturbed in the absence of any obvious glial disturbance. In other words, the network assumed a different responsiveness and the glial environment has adjusted to this new state. Clinically, this seems like a very important finding.

Of course, the problem with any negative finding is that there might be changes that the authors did not discover using the methods applied.

Reviewer #2:

This manuscript describes experiments designed to determine the effect of 1 time exposure to LPS during neonatal period in newborn rats on respiratory motor plasticity in adulthood. The work comes from a well-established laboratory with significant contributions to our understanding of phrenic nerve plasticity measured as long-term facilitation after several short exposures to intermittent hypoxia. They have outlined the 2 distinct cellular and molecular pathways that underlies this plasticity (S and Q pathway). They describe in this manuscript the effect of sex on the outcome variables.

Introduction: The Introduction nicely outlines the rationale for the experiments and points out that little is known about the effect of sex on expression of pLTF after neonatal exposure to inflammation. The topic is important since there are several adult onset diseases that are likely secondary to early neonatal exposure to inflammation such as Parkinsonism, and pLTF is thought to be a protective reflex to sustain ventilation during hypoxic exposure. The hypothesis is clearly stated that neonatal inflammation undermines the Q-pathway but not the S-pathway, respiratory motor plasticity in adult male and female rats.

The major findings are that LPS exposure in neonatal rats (1) increased mortality only in male pup, (2) abolished the phrenic nerve plasticity (pLTF) induced by moderate intermittent hypoxia involving the Q-pathway, and acute exposure to ketoprofen given to adults restored the pLTF, (3) similarly, the S-pathway mediating pLTF was affected in both male and female adult rats but ketoprofen did not restore the response, lastly (4) adult females had a greater acute phrenic amplitude response during moderate and severe hypoxia but this was unaffected by LPS exposure or Keto. Of note, markers of inflammation (cytokine gene expression markers of microglia activation) examined was not seen in the areas of the brainstem and cervical spinal cord). The authors conclude the effects observed in adult animals suggest neonatal inflammation affects peripheral inflammatory signaling or protein expression that mediates the responses.

Overall the findings are interesting and the paper well written but very long. I do have some concerns about the conclusion stated by the authors based on the experimental design and the data presented. The following questions should be addressed.

1) The authors use the term "motor plasticity" but the experiments only measured neural activity from the phrenic nerve. They also refer to "ventilator responses" the animals are anesthestized, paralyzed and vagotimzed, they don't have a "ventilatory response". The authors should simply state what was measured the electrical responses of the phrenic neurogram.

2) Is it possible that these experiments could have been done in awake animals? Can you observe the long-term motor plasticity and true ventilatory responses? Experiments in awake animals would be more meaningful. How does the phrenic responses translate into the true response that we are interested in -Ventilation?

3) What were serum glucose levels during the experimental challenges?

4) Why did the male animals have greater mortality after LPS exposure?

5) Was a dose response of LPS done in preliminary experiments? If not, why was the current dose of LPS chosen?

6) Did LPS exposure ever elicit an inflammatory response within the brains of these animals during the neonatal period? This would have been of interest to even know that the amount given ever elicited a central response.

7) What "peripheral" systems do the authors believe the early LPS exposure is modifying.

8) Why did the authors choose ketoprofen vs other non-steroidal anti-inflammatory drugs or even steroids?

9) The authors also conclude that their data indicate that a persistent change in the adult inflammatory signaling contributes to the over findings since Ketoprofen restores the response. Does this imply that the S pathway is not the modified by inflammation? The authors never demonstrate that "any inflammation" (central or peripheral) exist in the animals What are the off target non-anti-inflammatory effects of ketoprofen? Is the difference between the effects of the ketoprofen on the Q and S pathway simply because a high enough dose wasn't used?

10) Tissues were examined after the acute experiment was performed. Were tissues from LPS and saline exposed animals examined prior to the acute hypoxic experiments? Is it possible that the acute exposures may have modified the gene and protein expression pattern?

11) The paper is quite long, and a lot of the findings show no difference between the groups; I think the figures could be reduced by 25%. There is also an excessive number of references.

12) Not sure that is necessary for the authors to reference so many papers from their laboratory. There are an excessive number of references.

Reviewer #3:

The study "A single bout of neonatal inflammation impairs adult respiratory motor plasticity in male and female rats" by Hocker et al., shows that a single application of Lipopolysaccharide (LPS) after birth produces a reduction of different forms of long-term facilitation (LTF), which can be reverted by ketoprofen, in some cases. The finding that early inflammation induces long-lasting effects on breathing plasticity agrees with a wide variety of findings showing that early inflammation induces long-lasting effects on a several brain circuits (Walker et al., 2003; Spencer et al., 2006; Li et al., 2010; Kentner et al., 2010; Kirsten et al., 2011; Fan et al., 2011; Wang et al., 2011; Cai et al., 2013; Wang et al., 2013; Rideau Batista Novais et al., 2014; Lan et al., 2015; Comim et al., 2016; Patil et al., 2016; Onufriev et al., 2017; Singh et al., 2017; Réus et al., 2017; including the spinal cord; Walker et al., 2003; Li et al., 2010) and their plasticity (Walker et al., 2003; Rideau Batista Novais et al., 2014; Onufriev et al., 2017). Surprisingly, in contrast with other reports, the effects observed in this study did not correlate with changes in glia morphology (for contrast see: Fan et al., 2011; Wang et al., 2011; Cai et al., 2013; Wang et al., 2013; Yin et al., 2013; Pang et al., 2013; Lan et al., 2015) and the expression of pro-inflammatory genes (for contrast see: Cai et al., 2013; Yin et al., 2013; Patil et al., 2016, Onufriev et al., 2017) or proinflammatory mediators (for contrast see: Spencer et al., 2006; Kentner et al., 2010; Fan et al., 2011; Wang et al., 2013; Cai et al., 2013; Pang et al., 2013; Lunardelli et al., 2015; Lan et al., 2015; Patil et al., 2016; Onufriev et al., 2017; Réus et al., 2017). I consider the findings to be relevant, but also partially similar to the already described changes in other circuits of the brain by the same experimental manipulation (Walker et al., 2003;Spencer et al., 2006; Li et al., 2010; Kentner et al., 2010; Kirsten et al., 2011; Fan et al., 2011; Wang et al., 2011; Cai et al., 2013; Wang et al., 2013; Rideau Batista Novais et al., 2014; Lan et al., 2015; Comim et al., 2016; Patil et al., 2016; Onufriev et al., 2017; Singh et al., 2017; Réus et al., 2017). On the other hand, the mechanisms behind the findings of the present study are still foggy, since it is not clear what is the source of the inflammatory and the non-inflammatory signals involved in the described effects. Considering the contradictory evidence already available (Walker et al., 2003;Spencer et al., 2006; Li et al., 2010; Kentner et al., 2010; Kirsten et al., 2011; Fan et al., 2011; Wang et al., 2011; Cai et al., 2013; Wang et al., 2013; Rideau Batista Novais et al., 2014; Lan et al., 2015; Comim et al., 2016; Patil et al., 2016; Onufriev et al., 2017; Singh et al., 2017; Réus et al., 2017), the clarification of the mechanisms involved in the modulation described in this work would contribute to the relevance of the study.

[Editors’ note: what now follows is the decision letter after the authors submitted for further consideration.]

Thank you for submitting your article "One bout of neonatal inflammation impairs adult respiratory motor plasticity in male and female rats" for consideration by eLife. Your article has been reviewed by three peer reviewers, including Jan-Marino Ramirez as the Reviewing Editor and Reviewer #1 and the evaluation has been overseen by a Reviewing Editor and Ronald Calabrese as the Senior Editor.

The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

Summary:

The manuscript by Hocker et al., has important implications. Many babies are exposed to inflammatory events early during their development, and due to an improved healthcare, these babies mostly survive and recover. Yet, it is unclear whether this recovery is complete, and whether even in the absence of any acute inflammation there are long-term consequences on respiratory control. The present study addresses this important issue employing an established rodent model that uses respiratory plasticity in the phrenic nerve as a well-defined respiratory output.

This manuscript is a resubmission of a previously reviewed manuscript, and the authors performed additional experiments to address several of the reviewers' original comments. Their new experiments shed new light on the potential involvement of astrocytes (as it relates to the S-pathway in case of severe AIH). The authors also added new experimental data indicating distinct differences between the impairment of Q and S-pathways caused by neonatal inflammation. These are major revisions to the previous manuscript – which is particularly remarkable since such developmental studies consume considerable time.

The authors also added a discussion of the importance of acute intermittent hypoxia in the context of spinal cord injury and the therapeutic application for motor neuron disease. The authors emphasize the important implication in particular that the application of AIH could be undermined by early inflammatory experiences.

The authors also included a table from a separate study that describes in detail the inflammatory gene data that are associated with the LPS induced inflammation. However, it makes sense that these data are not included in the present study – because they are published elsewhere. However, these data were helpful for evaluating the findings of the present study, because their careful characterization confirms that there is impairment of respiratory activity even though the inflammatory gene responses have returned to baseline.

This leads to the important conclusion that the observed loss of plasticity is unlikely due to an ongoing inflammation.

Essential revisions:

The reviewers concluded that no additional experiments will be required. However, some concerns remained unaddressed, and need to be carefully addressed in the revision.

1) The authors should include a limitation section that discusses the experimental approach: in particular whether their findings can be translated to a fully integrated/ intact model animal or human. Regardless of how the model might allow one to dissect out pathways that underlie pLTF, if the whole purpose of the reflex is to maintain ventilation to compensate for hypoxia, then it is still important to demonstrate that the findings are relevant to ventilation in intact, unanesthetized animals. Please address what the findings mean behaviorally, if they are transmitted to the muscles of respiration?

2) We suggest reducing the number of figures with negative findings, and instead include a table showing what is different between the groups, which pathway is implicated, and the gender effects. The table would allow the reader to get a better conceptual view of all the many variables that are being assessed in this study.

3) The Abstract is very short and not very informative. The age of the animal at the time of the exposure and the age of the animals when the pLTF was assessed is not included. Moreover, there are sex differences that also are not specified in the Abstract.

4) In their response, the authors have clearly indicated the novelty of their relevant findings, in contrast to the wealth of evidence regarding the long-term consequences of early-life inflammation. Which, by the way, was clear in the original submission. However, we think that it is still necessary to carefully deal with the fact that among those published effects there is a long-lasting change in astrocytes and microglia as well as in their inflammatory mediators. Since these changes were not observed in the submitted study, the authors should include a proper discussion of the source of these differences and they should discuss the alternative mechanisms that could potentially be involved. The authors have shown several aspects that are not involved in their novel findings, which we agree is very important. The authors have also included a set of experiments that indirectly indicate that the astrocytic function could be affected by early-life inflammation, but this needs to be carefully discussed. We agree with the authors that although indirect, these experiments hint that astrocytes could have been changed by early-life inflammation. What we would like to see in the revision is that the authors discuss future experiments or avenues that could more directly address their observations.

https://doi.org/10.7554/eLife.45399.025

Author response

[Editors’ note: a previous version of this study was rejected after peer review, but the authors submitted for reconsideration. The first decision letter after peer review is shown below.]

Our decision has been reached after consultation between the reviewers. Based on these discussions and the individual reviews below, we regret to inform you that your work will not be considered further for publication in eLife.

We thank the reviewers for their recommendations, and have completed additional experiments to highlight distinct, mechanistic differences between Q- and S-pathway impairment after neonatal inflammation. We also now include additional evidence demonstrating that phrenic motor neurons are not impaired after neonatal inflammation, suggesting impairment in glial cell (e.g. astrocyte) function. Given the developmental nature of this study, collection and analysis of new data represent significant changes in the manuscript from the original submission and significantly enhance the overall impact of the manuscript. Lastly, we modified the manuscript to emphasize the clinical relevance of these studies by discussing the use of acute intermittent hypoxia (AIH)-induced plasticity therapeutically after spinal cord injury, and potential use for treatment of motor neuron disease. Our novel findings demonstrate that such therapeutic approaches would be undermined by experiences early in life.

The reviewers felt that additional experiments will be necessary to increase the significance of your findings. The reviewers had several specific recommendations and suggestions.

1) It would be important to demonstrate that there is indeed significant inflammation in areas relevant for LTF – e.g. phrenic nucleus – 48hours after LPS injection and 1 or two weeks later.

Thank you for raising this issue, as we are also interested in the inflammatory state of CNS regions relevant to respiratory plasticity after LPS. The ability of LPS to induce inflammation at a variety of doses are well documented both in neonates and adults in many CNS regions (reviewed in Bilbo and Schwarz, 2012). Indeed, we have gone to considerable effort to characterize the effects of LPS-induced inflammation in respiratory control regions after neonatal LPS. In the context of this manuscript, the most salient point is that in adults, long after the inflammatory gene response has returned to baseline levels after neonatal LPS challenge, there are lasting impairments in respiratory activity. Figure 4 in the manuscript shows that there is no significant difference in inflammatory gene expression in adults challenged with either neonatal saline or LPS, despite the loss of respiratory motor plasticity in the LPS-challenged adults. Thus, while neonatal inflammation causes transient increases in the expression of inflammatory genes, the lasting effects of neonatal inflammation into adulthood either result from cell-type specific inflammatory signaling changes not captured in our analysis, or ongoing inflammation in other CNS regions impacting spinal motor plasticity.

2) The reviewers also recommended to add some experiments that show that LTF translates into an altered hypoxic ventilatory response in freely moving unanesthetized animal.

We agree that changes in ventilatory responses in freely behaving animals are of interest, and those studies represent the next stage of experiments in this project. The present study is the first to characterize lasting impairments in the neural control of breathing after a single inflammatory insult during the neonatal period, and our current electrophysiologic approach has important strengths not afforded in freely behaving animals: (1) it enables measurement of changes in neural control independent of the mechanics of breathing, and (2) it allows us to better investigate mechanisms of respiratory motor plasticity using direct application of drugs onto the cervical spinal cord, where LTF occurs (see response to #3 above). While it is possible to measure respiratory motor plasticity in freely behaving animals (e.g. ventilatory LTF, see response to reviewer 2 #2), our interest is in the change in neural output caused by neonatal inflammation. Neuroplasticity cannot be studied in the context of whole animal breathing, and the powerful electrophysiologic model we have used here has allowed the identification of several cellular mechanisms underlying AIH-induced pLTF (reviewed in Turner et al., 2018). Further, studying spinal motor plasticity induced by acute intermittent hypoxia is of particular clinical relevance because intermittent hypoxia not only enhances respiratory motor output, but it also facilitates motor output throughout the spinal cord, and is used therapeutically to improve locomotor function in patients after spinal cord injury (Trumbower et al., 2012, 2017; Hayes et al., 2014). Thus, the impact of early life inflammation on motor neuron plasticity has clinical relevance for the utility of acute intermittent hypoxia as a therapeutic in regions independent of ventilation, underscoring the importance of studying plasticity in its own right. The restoration of spinal motor plasticity with acute anti-inflammatories suggests that anti-inflammatories may enhance the therapeutic benefit of acute intermittent hypoxia on motor output. We have revised the discussion accordingly, to include this important possibility (Discussion section).

3) Another issue that was raised relates to the apparent lack of a glial involvement. It would be more convincing to the readers, if the authors would add a deeper glial characterization e.g. trough flow cytometry, an approach which you seem to be familiar with (Nikodemova et al., 2012; 2014; 2015; 2016), and/or an additional test to explore whether glial responsivity to inflammatory challenges has been altered (Li et al., 2010; Cai et al., 2013; Yin et al., 2013; Lunardelli et al., 2015). It is possible that the glia seem unchanged under "resting" conditions but become more prone (primed) towards activation, something you are also very familiar with (Huxtable et al., 2013; Johnson et al., 2018).

We agree that it is possible that glia appear morphologically normal in adults, but that they are “primed” to respond differently after a subsequent adult stimulus. It is however important to highlight that the deficits in respiratory control presented here are independent of later life inflammatory events – findings that contrast with other studies showing the physiological effects of early life inflammation are only revealed after a subsequent adult stimulus (Bilbo and Schwarz, 2009; Wang et al., 2010; Bilbo, 2010; Kirsten et al., 2010, 2011; Roughton et al., 2013). That respiratory plasticity is impaired in the absence of a subsequent challenge underscores the novelty of our study and emphasizes the profound sensitivity of the neonatal respiratory circuitry to a single bout of early-life inflammation.

Nonetheless, based on the reviewer comments, we have now added an additional set of experiments that advance our understanding of the mechanistic underpinnings of, and glial involvement in, the impairment of S-pathway-evoked pLTF. The S-pathway is adenosine dependent (Nichols et al., 2012) and astrocytes are the primary source of adenosine during hypoxia (Takahashi et al., 2010; Angelova et al., 2015). Thus, we tested the hypothesis that astrocyte-phrenic motor neuron signaling is impaired in adults after neonatal inflammation. To test this idea, we utilized what is known about the mechanisms of S-pathway-evoked plasticity (Golder et al., 2008; Nichols et al., 2012), and episodically applied an adenosine receptor agonist. Our new data demonstrate that phrenic motor neurons are not generally impaired after neonatal inflammation as adenosine was able to elicit S-pathway-evoked pLTF. Rather, these new data suggest that the lasting impairment in S-pathway-evoked pLTF may be due to an alteration in adult spinal astrocytes, since the likely source of adenosine during severe intermittent hypoxia (sAIH) is astrocytes (Takahashi et al., 2010; Angelova et al., 2015). These results do not, however, rule out a role for altered microglia-astrocyte-neuron signaling, and the role of glia continue to be a topic of ongoing studies in our group. We are presently pursuing flow cytometry and RNAseq to better understand the status of astrocytes (and microglia) in adults following neonatal inflammation; these complex molecular analyses are beyond the scope of the present study.

Reviewer #1:

This is a solid study by Hocker et al., investigating the consequences of neonatal LPS injection on respiratory plasticity. The authors report several interesting findings.

One finding is the increased neonatal mortality observed in male versus female pups. This is a very important finding in the context of SIDS, which affects males significantly more than female. The ratio of mortality seen in this study is similar to that seen in human SIDS. The only minor concern is that this interesting observation is not further explored. Hence it seems a bit of an "add-on" rather than an integral part of the story reported in this manuscript.

Thank you for pointing out this important omission. We have now added the sentence below to our Discussion section to further highlight the importance of this finding. We are very interested in the sex differences in adult respiratory plasticity after neonatal inflammation, and studies are planned to further dissect their underlying mechanisms. Please also refer to response to reviewer 2 #4.

“Neonatal inflammation also increases male mortality consistent with clinical male mortality after neonatal inflammation (Person et al., 2014) and relevant to the increased risk of sudden infant death syndrome for males (Kinney and Thach, 2009).”

The remainder of the manuscript focuses on respiratory plasticity. It follows in the footsteps of the Gordon Mitchell laboratory and carefully dissects the so called "Long-term facilitation" into the different cellular pathways and the author characterize the underlying mechanisms of the LTF as well as the changes in female and males exposed to neonatal LPS.

The single bolus of LPS has long-term differential consequences on this form of plasticity and the underlying mechanisms. The authors perform e.g. a careful analysis of the cytokines involved – and the involvement is rather diverse. This is perhaps not surprising.

The one finding that seems unexpected is that the authors cannot identify the actual mechanism underlying the long-term impairment of the plasticity. They did not discover any changes in astrocytes and microglia, neither abundance of these cells nor in their morphology. This "negative finding" leaves the reader somewhat "unsatisfied" as it makes the study appear somewhat descriptive and "unresolved". However, this may not be the case – the authors have done a great job in unraveling the subcellular mechanisms underlying the plasticity and the changes associated with the one-time inflammatory event. Perhaps the lack of glial alterations is the most important finding, i.e. that the plasticity is disturbed in the absence of any obvious glial disturbance. In other words, the network assumed a different responsiveness and the glial environment has adjusted to this new state. Clinically, this seems like a very important finding.

Of course, the problem with any negative finding is that there might be changes that the authors did not discover using the methods applied.

The authors thank the reviewer for their positive comments and careful consideration of the importance of a “negative finding”. We agree that the results are both intriguing and surprising (and also perhaps a bit frustrating!). However, as with any complex physiological process, it can be challenging at first to unravel all of the mechanisms underlying a phenomenon. As described above (response to #3 above), we have undertaken further studies to better understand the mechanisms underlying abolition of plasticity. Our results indicate adult impairment of the Q-pathway is a result of an inflammation-dependent process, and we are performing a number of molecular analyses to identify the molecule(s) key to this impairment. However, the inflammatory response is complex, and identifying the “culprit” inflammatory molecule will require extensive molecular analyses. Such studies represent the next several years of our research plan, and are thus, beyond the scope of the current manuscript.

Impairment of the S-pathway, on the other hand, appears to be independent of inflammatory signaling, and our new data (Figure 7 in the manuscript) demonstrate phrenic motor neurons are not impaired after neonatal inflammation. We propose that S-pathway impairment is due to a lasting impairment in adenosine signaling, the source of which (astrocytes) is likely altered by neonatal inflammation. Intermittent application of adenosine induces S-pathway plasticity by activating adenosine 2A receptors on phrenic motor neurons (Seven et al., 2017), a pathway that appears normal after neonatal inflammation, suggesting that phrenic motor neurons themselves do not exhibit lasting detriments in motor output after neonatal inflammation. We have restructured the order of the results to clarify the mechanistic results in this manuscript.

Reviewer #2:

This manuscript describes experiments designed to determine the effect of 1 time exposure to LPS during neonatal period in newborn rats on respiratory motor plasticity in adulthood. The work comes from a well-established laboratory with significant contributions to our understanding of phrenic nerve plasticity measured as long-term facilitation after several short exposures to intermittent hypoxia. They have outlined the 2 distinct cellular and molecular pathways that underlies this plasticity (S and Q pathway). They describe in this manuscript the effect of sex on the outcome variables.

Introduction: The Introduction nicely outlines the rationale for the experiments and points out that little is known about the effect of sex on expression of pLTF after neonatal exposure to inflammation. The topic is important since there are several adult onset diseases that are likely secondary to early neonatal exposure to inflammation such as Parkinsonism, and pLTF is thought to be a protective reflex to sustain ventilation during hypoxic exposure. The hypothesis is clearly stated that neonatal inflammation undermines the Q-pathway but not the S-pathway, respiratory motor plasticity in adult male and female rats.

The major findings are that LPS exposure in neonatal rats (1) increased mortality only in male pup, (2) abolished the phrenic nerve plasticity (pLTF) induced by moderate intermittent hypoxia involving the Q-pathway, and acute exposure to ketoprofen given to adults restored the pLTF, (3) similarly, the S-pathway mediating pLTF was affected in both male and female adult rats but ketoprofen did not restore the response, lastly (4) adult females had a greater acute phrenic amplitude response during moderate and severe hypoxia but this was unaffected by LPS exposure or Keto. Of note, markers of inflammation (cytokine gene expression markers of microglia activation) examined was not seen in the areas of the brainstem and cervical spinal cord). The authors conclude the effects observed in adult animals suggest neonatal inflammation affects peripheral inflammatory signaling or protein expression that mediates the responses.

Overall the findings are interesting and the paper well written but very long. I do have some concerns about the conclusion stated by the authors based on the experimental design and the data presented. The following questions should be addressed.

1) The authors use the term "motor plasticity" but the experiments only measured neural activity from the phrenic nerve. They also refer to "ventilator responses" the animals are anesthestized, paralyzed and vagotimzed, they don't have a "ventilatory response". The authors should simply state what was measured the electrical responses of the phrenic neurogram.

Thank you for this comment and correction. Yes, we do use “motor plasticity” to describe LTF, as is common practice in studies using this model (Tadjalli et al., 2010; Agosto-Marlin et al., 2017; Lui et al., 2018; Seven et al., 2018). The manuscript has now been revised and the reference to ventilatory responses has been removed.

2) Is it possible that these experiments could have been done in awake animals? Can you observe the long-term motor plasticity and true ventilatory responses? Experiments in awake animals would be more meaningful. How does the phrenic responses translate into the true response that we are interested in -Ventilation?

The reviewer is correct in that aspects of the experiments here could be conducted in awake, freely behaving animals. For example, ventilatory LTF (vLTF), a lasting increase in ventilation after intermittent hypoxia in awake animals, can be used to monitor neuroplastic changes in breathing. However, a number of important limitations hinder the utility of using vLTF as a model for understanding respiratory neuroplasticity; the magnitude of vLTF is small (Olson et al., 2001), sleep state rapidly alters vLTF (Nakamura et al., 2010), and vLTF is much more variable than phrenic LTF (Olson et al., 2001). Furthermore, pLTF is a more useful model for understanding mechanistic aspects of respiratory plasticity, as the mechanistic underpinnings of pLTF have been studied by us and others for more than 25 years (Turner et al., 2018). Thus, we chose the most commonly studied model of respiratory plasticity, pLTF, to further probe mechanisms undermining adult plasticity after neonatal inflammation (see also response to #2 above). As exemplified in our newly included data (Figure 7) using this model, we demonstrate there is likely an astrocytic, adenosine-dependent mechanism disrupted by neonatal inflammation, which is responsible for undermining adult plasticity.

3) What were serum glucose levels during the experimental challenges?

In recent experiments, we have measured blood glucose levels at the beginning and end of phrenic nerve recordings. Due to the use of the anesthetic urethane (Wang et al., 2000), blood glucose levels are elevated in all adults regardless of whether they received neonatal saline or LPS. Importantly, adult blood glucose levels do not differ between adults that received neonatal saline or LPS (p = 0.2739), and they do not change over the course of electrophysiology recordings.

P4 Saline = 225 +/- 38 mg/dL, n=6 (4 males, 2 females) P4 LPS = 251 +/- 27 mg/dL, n=4 (2 males, 2 females)

4) Why did the male animals have greater mortality after LPS exposure?

The reviewer highlights an interesting aspect of our current study. The increase in male mortality is consistent with other reports of increased male vulnerability early in life. Interestingly, sex differences exist early in brain development to either establish or eliminate sex differences in brain function and behavior. Sex differences in microglial colonization exist at P4 in the hippocampus, parietal cortex, and amygdala with male rats having significantly more microglia than females (Bilbo and Schwarz, 2012; Schwarz et al., 2012). These sex differences have been proposed to explain different susceptibilities to certain developmental psychiatric disorders (Bilbo and Schwarz, 2012; Nelson and Guyer, 2012; Schwarz et al., 2012). Whether such differences also exist in respiratory-related regions are unknown but are currently under investigation in our laboratory. We chose not to speculate on the reason(s) for this increased mortality in males until we had more data to support such suggestions.

Although we failed to see any persistent differences in glial number or gross morphology in our current experiments, we and others, have hypothesized persistent differences in glial activities, and thus their behavior, might underlie increased male vulnerability (Nelson and Guyer, 2012; Reemst et al., 2016). Our understanding of neonatal astrocyte distribution and migration remains incomplete, but consensus is that astrogenesis continues postnatally. Both astrocytes and microglia have important roles in neuronal survival and programmed cell death. Synaptogenesis peaks postnatally and corresponds with astrogenesis. Further, glial dysfunction has been linked to numerous neurodevelopmental disorders, including: autism spectrum disorder (Zhan et al., 2014; Squarzoni et al., 2014), schizophrenia (Ma et al., 2013; Zhan et al., 2014), Fragile-X syndrome (Jacobs et al., 2010; Higashimori et al., 2013), depression (Nishiyama et al., 2002), and obsessive compulsive disorder (Zhan et al., 2014). Glial dysfunction has also been linked to Rett syndrome (Maezawa et al., 2009), which manifests with respiratory deficits such as irregular breathing, hyperventilation and apneas (Katz et al., 2009; Ren et al., 2012; Ramirez et al., 2013). Thus, because we are only in the beginning stages of characterizing sex differences in neural development which may underlie increased male mortality, this remains an area of active investigation in our laboratory and others’. Should the reviewers feel this warrants additional discussion in the manuscript, we are happy to include it; but we have remained cognizant of increasing the length of the manuscript given that the reviewers’ felt it was already too long.

5) Was a dose response of LPS done in preliminary experiments? If not, why was the current dose of LPS chosen?

We chose 1mg/kg LPS in our study based on our unpublished dose-response experiments. We found this to be the best dose where both neonatal mortality was minimal and inflammatory gene expression was transiently increased in both the neonatal medulla and cervical spinal cord, key respiratory control regions (see response 1 above for gene expression data). Our observations are also in line with inflammatory gene expression in other reports in neonates in both the medulla and spinal cord after 1 mg/kg LPS (Wang et al., 2006; Balan et al., 2011; Schwarz and Bilbo, 2011; Jafri et al., 2013), and are consistent with other experiments showing that higher LPS doses induce significant neonatal mortality (Blood-Siegfried et al., 2002; Rourke et al., 2016).

6) Did LPS exposure ever elicit an inflammatory response within the brains of these animals during the neonatal period? This would have been of interest to even know that the amount given ever elicited a central response.

Yes, LPS did initiate CNS inflammation in the neonate, and specifically in the medulla and cervical spinal cord (see response #1 above).

7) What "peripheral" systems do the authors believe the early LPS exposure is modifying.

We regret the lack of clarity on this point, and have revised the discussion accordingly. Since the anti-inflammatory ketoprofen restored Q-pathway-evoked plasticity even in the absence of detectable CNS inflammation in adults, our results suggest inflammatory signaling (either central and/or peripheral) contributes to plasticity deficits in adults after neonatal inflammation. We hypothesize that other inflammatory molecules not evaluated here, or inflammation in other important brain regions or peripheral tissues may be responsible for the lasting ketoprofen-sensitive inflammatory signaling undermining adult Q-pathway-evoked plasticity. RNA-Seq analyses are in progress to evaluate all inflammatory genes expressed.

“Additionally, no obvious differences in astrocyte or microglial morphology in adult phrenic motor nuclei or the preBötC were seen following neonatal LPS inflammation, suggesting other inflammatory mechanisms may be responsible for impairing adult pLTF.” Results section.

8) Why did the authors choose ketoprofen vs other non-steroidal anti-inflammatory drugs or even steroids?

Ketoprofen has previously been used to restore respiratory motor plasticity after acute inflammatory stimuli in adult rats (Huxtable et al., 2013, Huxtable et al., 2015). Ketoprofen is a non-selective NSAID targeting both COX-1 and COX-2, as well as NF-κB at the high doses used here (Cashman, 1996, Yin et al., 1998). Since acute inflammatory impairment of LTF in adults is COX-independent (Huxtable et al., 2018), we suggest that ketoprofen restores LTF either by inhibiting NF-κB activation (and thus subsequent inflammatory gene activation) or by other COX-independent mechanisms. Glucocorticoids, on the other hand, are well-known to suppress BDNF expression, a molecule critical for Q-pathway-evoked pLTF. Thus, the use of steroidal anti-inflammatories would have precluded our ability to accurately interpret the results.

9) The authors also conclude that their data indicate that a persistent change in the adult inflammatory signaling contributes to the over findings since Ketoprofen restores the response. Does this imply that the S pathway is not the modified by inflammation? The authors never demonstrate that "any inflammation" (central or peripheral) exist in the animals What are the off target non-anti-inflammatory effects of ketoprofen? Is the difference between the effects of the ketoprofen on the Q and S pathway simply because a high enough dose wasn't used?

Please also see response to reviewer 2 #8 above. As discussed in our previous publications (Huxtable et al., 2015; Huxtable et al., 2018), ketoprofen is a potent and common anti-inflammatory and analgesic agent used in many species, including humans (Foster et al., 1988) and rats (Cabre et al., 1998). The dose of ketoprofen used here is nearly 8x the median effective dose in rats (Cabré et al., 1998), so based on drug solubility and toxicity studies in other reports, higher doses are not feasible. The acute inflammation-independent nature of the S-pathway has been previously reported (Agosto-Marlin et al., 2017), so it follows that blocking acute inflammatory signaling in adults after neonatal inflammation would not restore LTF. We do not think the differential sensitivity of the Q and S pathways to inflammation are dependent on the ketoprofen dose, but rather, the lasting deficit in S-pathway plasticity is due to an astrocytic, adenosine-dependent mechanism (see response to reviewer 1).

10) Tissues were examined after the acute experiment was performed. Were tissues from LPS and saline exposed animals examined prior to the acute hypoxic experiments? Is it possible that the acute exposures may have modified the gene and protein expression pattern?

Tissues taken for gene expression analysis were taken from separate animals not used for neurophysiology experiments. However, tissues for immunohistochemistry experiments were taken at the end of the acute neurophysiology experiment because previous experiments by our group have shown no significant differences in inflammatory gene expression after AIH in CNS tissues (Smith et al., 2013). This practice is routine in our lab to reduce the total number of animals used. The methods now correctly reflect how tissues were used.

11) The paper is quite long, and a lot of the findings show no difference between the groups; I think the figures could be reduced by 25%. There is also an excessive number of references.

We agree that there is a substantial amount of data in this manuscript. However, due to the complexity of groups (neonatal inflammation/saline, age, sex, and other treatments), we separated the figures based on the individual hypotheses being tested to reduce density in the figures and improve clarity. If the reviewer has suggestions regarding which figures could be combined or eliminated, we would be happy to re-evaluate.

12) Not sure that is necessary for the authors to reference so many papers from their laboratory. There are an excessive number of references.

The references to our previous papers appear most in the Introduction and Materials and methods sections and were cited because they laid the foundation for the present study by establishing the acute effects of inflammation in adults. Additionally, much of the methodology used here has previously been used by our group, so to reduce reiterating previously published work, we instead direct readers to the previously published work. However, we have now removed redundant references in the Introduction and Discussion section.

Reviewer #3:

The study "A single bout of neonatal inflammation impairs adult respiratory motor plasticity in male and female rats" by Hocker et al,. shows that a single application of Lipopolysaccharide (LPS) after birth produces a reduction of different forms of long-term facilitation (LTF), which can be reverted by ketoprofen, in some cases. The finding that early inflammation induces long-lasting effects on breathing plasticity agrees with a wide variety of findings showing that early inflammation induces long-lasting effects on a several brain circuits (Walker et al., 2003; Spencer et al., 2006; Li et al., 2010; Kentner et al., 2010; Kirsten et al., 2011; Fan et al., 2011; Wang et al., 2011; Cai et al., 2013; Wang et al., 2013; Rideau Batista Novais et al., 2014; Lan et al., 2015; Comim et al., 2016; Patil et al., 2016; Onufriev et al., 2017; Singh et al., 2017; Réus et al., 2017; including the spinal cord; Walker et al., 2003; Li et al., 2010) and their plasticity (Walker et al., 2003; Rideau Batista Novais et al., 2014; Onufriev et al., 2017). Surprisingly, in contrast with other reports, the effects observed in this study did not correlate with changes in glia morphology (for contrast see: Fan et al., 2011; Wang et al., 2011; Cai et al., 2013; Wang et al., 2013; Yin et al., 2013; Pang et al., 2013; Lan et al., 2015) and the expression of pro-inflammatory genes (for contrast see: Cai et al., 2013; Yin et al., 2013; Patil et al., 2016, Onufriev et al., 2017) or proinflammatory mediators (for contrast see: Spencer et al., 2006; Kentner et al., 2010; Fan et al., 2011; Wang et al., 2013; Cai et al., 2013; Pang et al., 2013; Lunardelli et al., 2015; Lan et al., 2015; Patil et al., 2016; Onufriev et al., 2017; Réus et al., 2017). I consider the findings to be relevant, but also partially similar to the already described changes in other circuits of the brain by the same experimental manipulation (Walker et al., 2003;Spencer et al., 2006; Li et al., 2010; Kentner et al., 2010; Kirsten et al., 2011; Fan et al., 2011; Wang et al., 2011; Cai et al., 2013; Wang et al., 2013; Rideau Batista Novais et al., 2014; Lan et al., 2015; Comim et al., 2016; Patil et al., 2016; Onufriev et al., 2017; Singh et al., 2017; Réus et al., 2017). On the other hand, the mechanisms behind the findings of the present study are still foggy, since it is not clear what is the source of the inflammatory and the non-inflammatory signals involved in the described effects. Considering the contradictory evidence already available (Walker et al., 2003;Spencer et al., 2006; Li et al., 2010; Kentner et al., 2010; Kirsten et al., 2011; Fan et al., 2011; Wang et al., 2011; Cai et al., 2013; Wang et al., 2013; Rideau Batista Novais et al., 2014; Lan et al., 2015; Comim et al., 2016; Patil et al., 2016; Onufriev et al., 2017; Singh et al., 2017; Réus et al., 2017), the clarification of the mechanisms involved in the modulation described in this work would contribute to the relevance of the study.

We thank the reviewer for their comments and the description of our work in the context of the neuroscience field. We agree that aspects of our results are similar to results in other CNS circuits, however, this is the first demonstration of such changes in a vital homeostatic circuit. Neural circuitry involved in respiratory control must adapt continuously to maintain homeostatic conditions, and plasticity is thought to play a key role. Failure of plasticity in this circuitry could have dire consequences, perhaps even explaining the increased mortality in males after neonatal LPS. While we are not making such broad speculations here, our results are novel and distinct in three ways:

1) Unlike much literature on early life events, we demonstrate profound lasting effects after just one neonatal stimulus. Other studies demonstrate alterations in adulthood after a second, adult stimulus (Bilbo and Schwarz, 2009; Wang et al., 2010; Bilbo, 2010; Kirsten et al., 2010, 2011; Roughton et al., 2013).

2) This is the first demonstration of abolition of S-pathway induced respiratory plasticity, which has been proposed to act as a “back-up pathway” to preserve plasticity in the face of inflammation. Thus, we hypothesize that early life inflammatory events represent the most significant perturbation of the respiratory plasticity. Further, we now provide additional mechanistic details about S-pathway abolishment by neonatal inflammation, and suggest astrocyte signaling is permanently impaired after neonatal inflammation (see response to #3).

3) While our data are consistent with increased male vulnerability early in life, we are the first to characterize female responses. Assessment of respiratory control in females and female respiratory motor plasticity fills a major gap in knowledge and directly addresses the NIH research priority for sex differences. The role of sex hormones in respiratory control remains in its infancy (Behan and Kinkead, 2011). Critical periods for sex hormone action exist during development; high testosterone in young males, and fluctuations of estrogen and progesterone in post-pubescent females are normal changes in hormones during development. Yet interactions between an environmental stimulus (like inflammation) and sex/hormonal differences during these critical periods are poorly understood. Neonatal maternal separation elicits a stress response with sexspecific effects and lasting consequences on the hypothalamic-pituitary-adrenal axis, baseline breathing, and chemosensivity, with males showing greater susceptibility (Genest et al., 2004).

Yet despite the longstanding “Barker Hypothesis” (developmental origins of adult disease) (Barker and Osmond, 1986; Barker, 2002) and other calls to explore the effects of neonatal events on respiratory control (Behan and Kinkead, 2011), our studies represent the first step to fill this gap in knowledge.

Determining how neonatal inflammation impairs the respiratory system will elucidate links between neonatal conditions and adult respiratory insufficiency, leading to better treatments to promote breathing at all age groups. Our study has implications for understanding these long-term deficiencies and will facilitate identification of location(s) within the complex respiratory neural circuitry this deficiency occurs. Previous studies (Rosen et al., 2003; Hibbs et al., 2008; Raynes-Greenow et al., 2012) were correlative, and provided important additions to the scientific premise for the present work; but this is the first to directly test the involvement of neonatal inflammation on respiratory circuitry and specific cell types in the CNS.

[Editors’ note: what now follows is the decision letter after the authors submitted for further consideration.]

Summary:

The manuscript by Hocker et al., has important implications. Many babies are exposed to inflammatory events early during their development, and due to an improved healthcare, these babies mostly survive and recover. Yet, it is unclear whether this recovery is complete, and whether even in the absence of any acute inflammation there are long-term consequences on respiratory control. The present study addresses this important issue employing an established rodent model that uses respiratory plasticity in the phrenic nerve as a well-defined respiratory output.

This manuscript is a resubmission of a previously reviewed manuscript, and the authors performed additional experiments to address several of the reviewers' original comments. Their new experiments shed new light on the potential involvement of astrocytes (as it relates to the S-pathway in case of severe AIH). The authors also added new experimental data indicating distinct differences between the impairment of Q and S-pathways caused by neonatal inflammation. These are major revisions to the previous manuscript – which is particularly remarkable since such developmental studies consume considerable time.

The authors also added a discussion of the importance of acute intermittent hypoxia in the context of spinal cord injury and the therapeutic application for motor neuron disease. The authors emphasize the important implication in particular that the application of AIH could be undermined by early inflammatory experiences.

The authors also included a table from a separate study that describes in detail the inflammatory gene data that are associated with the LPS induced inflammation. However, it makes sense that these data are not included in the present study – because they are published elsewhere. However, these data were helpful for evaluating the findings of the present study, because their careful characterization confirms that there is impairment of respiratory activity even though the inflammatory gene responses have returned to baseline.

This leads to the important conclusion that the observed loss of plasticity is unlikely due to an ongoing inflammation.

Essential revisions:

The reviewers concluded that no additional experiments will be required. However, some concerns remained unaddressed, and need to be carefully addressed in the revision.

1) The authors should include a limitation section that discusses the experimental approach: in particular whether their findings can be translated to a fully integrated/ intact model animal or human. Regardless of how the model might allow one to dissect out pathways that underlie pLTF, if the whole purpose of the reflex is to maintain ventilation to compensate for hypoxia, then it is still important to demonstrate that the findings are relevant to ventilation in intact, unanesthetized animals. Please address what the findings mean behaviorally, if they are transmitted to the muscles of respiration?

The authors thank the reviewers for their comments. We have now included a limitation section in the Discussion section addressing the potential impact to adult animals and humans. While we are still understanding the exact physiological role of respiratory plasticity, the ability for the respiratory system to adapt and learn is critical for maintaining homeostasis. Our results suggest at least two pathways associated with learning in the respiratory system are impaired following just one bout of early life inflammation. This suggests the respiratory system is likely more vulnerable in adulthood. We have added a more thorough discussion of the physiological implications and limitations of our findings:

“Our experimental approach assessed phrenic nerve output in anesthetized rats and may not be generalizable to respiratory control in freely behaving animals or to other forms of motor plasticity. In humans, AIH induces long-term facilitation of ventilation (Mateika and Komnenov, 2017) and strengthens corticospinal pathways to non-respiratory motor-neurons (Christiansen et al., 2018), suggesting our results likely have relevance to mechanisms of human spinal motor plasticity after AIH. While AIH-induced respiratory motor plasticity does not necessarily alter normal homeostatic control of ventilation, the general facilitation of spinal motor output has significant therapeutic potential for treating patients with respiratory and non-respiratory motor limitations (Trumbower et al., 2012, 2017; Nichols et al., 2013; Hayes et al., 2014).”

2) We suggest reducing the number of figures with negative findings, and instead include a table showing what is different between the groups, which pathway is implicated, and the gender effects. The table would allow the reader to get a better conceptual view of all the many variables that are being assessed in this study.

To improve the clarity of our results, we have reorganized and consolidated figures. As we did not find sex differences in Q-pathway or S-pathway evoked motor plasticity, and the only sex difference in response to neonatal inflammation was the acute neonatal male mortality, we feel a table is not necessary. We have removed two figures and reorganized the Results section to further streamline the manuscript. Figure 9 and Figure 10 showing no changes in preBötC and cervical spinal glial morphology have been combined into one figure (now Figure 8). The previous Figure 8, showing phrenic nerve responses to hypoxia, has been converted to a table (Table 1) and moved to the end of the Results section. Additionally, we have combined male and female inflammatory gene expression data since no sex differences in respiratory plasticity were evident and to reduce the number of figures showing negative data.

3) The Abstract is very short and not very informative. The age of the animal at the time of the exposure and the age of the animals when the pLTF was assessed is not included. Moreover, there are sex differences that also are not specified in the Abstract.

The authors thank the reviewers for their comments. We have revised the Abstract to include additional information; however, we are confined by the eLife limit of 150 words.

No sex differences were found in any primary outcomes mentioned in the abstract. The sex-differences in neonatal mortality after acute LPS exposure is a relevant finding and may importantly relate to male-specific human mortality; however, the abstract length limitation does not allow adequate discussion of this finding. Our revised Abstract is below.

“Neonatal inflammation is common and has lasting consequences for adult health. We investigated the lasting effects of a single bout of neonatal inflammation on adult respiratory control in the form of respiratory motor plasticity induced by acute intermittent hypoxia, which likely compensates and stabilizes breathing during injury or disease and has significant therapeutic potential. Lipopolysaccharide-induced inflammation at postnatal day four induced lasting impairments in two distinct pathways to adult respiratory plasticity in male and female rats. Despite a lack of adult pro-inflammatory gene expression or alterations in glial morphology, one mechanistic pathway to plasticity was restored by acute, adult anti-inflammatory treatment, suggesting ongoing inflammatory signaling after neonatal inflammation. An alternative pathway to plasticity was not restored by anti-inflammatory treatment, but was evoked by exogenous adenosine receptor agonism, suggesting upstream impairment, likely astrocytic-dependent. Thus, the respiratory control network is vulnerable to early-life inflammation, limiting respiratory compensation to adult disease or injury.”

4) In their response, the authors have clearly indicated the novelty of their relevant findings, in contrast to the wealth of evidence regarding the long-term consequences of early-life inflammation. Which, by the way, was clear in the original submission. However, we think that it is still necessary to carefully deal with the fact that among those published effects there is a long-lasting change in astrocytes and microglia as well as in their inflammatory mediators. Since these changes were not observed in the submitted study, the authors should include a proper discussion of the source of these differences and they should discuss the alternative mechanisms that could potentially be involved. The authors have shown several aspects that are not involved in their novel findings, which we agree is very important. The authors have also included a set of experiments that indirectly indicate that the astrocytic function could be affected by early-life inflammation, but this needs to be carefully discussed. We agree with the authors that although indirect, these experiments hint that astrocytes could have been changed by early-life inflammation. What we would like to see in the revision is that the authors discuss future experiments or avenues that could more directly address their observations.

The authors thank the reviewers for their comments and interest in our study. Similar to other findings, we found no changes in adult inflammatory gene expression after neonatal LPS-induced inflammation. However, a few reports (Boisse et al., 2005; Kentner et al., 2010) have identified region and sex specific changes in adult CNS inflammation, which have been added to the Discussion section. Other perinatal stimuli with associated inflammatory signaling, such as maternal care and maternal diet, do have lasting programming effects, but are more complex stimuli than the neonatal inflammation in this study. Neonatal inflammation alone does prime glial responses to adult stimuli (Burke et al., 2016), but does not alter baseline inflammatory gene expression, consistent with the findings presented here. We have added additional detail on this to the Discussion section. Further, we also now include mention of our ongoing studies examining cell-specific changes to adult microglia and astrocytes, which we hypothesize will ultimately identify how neonatal inflammation impairs adult, respiratory plasticity (Discussion section).

https://doi.org/10.7554/eLife.45399.026

Article and author information

Author details

  1. Austin D Hocker

    Department of Human Physiology, University of Oregon, Eugene, United States
    Contribution
    Conceptualization, Data curation, Formal analysis, Investigation, Visualization, Methodology, Writing—original draft, Project administration, Writing—review and editing
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-2941-2581
  2. Sarah A Beyeler

    Department of Human Physiology, University of Oregon, Eugene, United States
    Contribution
    Investigation, Visualization, Writing—original draft
    Competing interests
    No competing interests declared
  3. Alyssa N Gardner

    Department of Comparative Biosciences, University of Wisconsin-Madison, Madison, United States
    Contribution
    Formal analysis, Investigation
    Competing interests
    No competing interests declared
  4. Stephen M Johnson

    Department of Comparative Biosciences, University of Wisconsin-Madison, Madison, United States
    Contribution
    Conceptualization, Resources, Investigation, Writing—review and editing
    Competing interests
    No competing interests declared
  5. Jyoti J Watters

    Department of Comparative Biosciences, University of Wisconsin-Madison, Madison, United States
    Contribution
    Conceptualization, Resources, Data curation, Funding acquisition, Project administration, Writing—review and editing
    Competing interests
    No competing interests declared
  6. Adrianne G Huxtable

    Department of Human Physiology, University of Oregon, Eugene, United States
    Contribution
    Conceptualization, Resources, Funding acquisition, Investigation, Methodology, Writing—original draft, Project administration, Writing—review and editing
    For correspondence
    huxtable@uoregon.edu
    Competing interests
    No competing interests declared
    ORCID icon "This ORCID iD identifies the author of this article:" 0000-0002-8745-2231

Funding

National Institutes of Health (HL141249)

  • Adrianne G Huxtable

National Institutes of Health (HL111598)

  • Jyoti J Watters

The funders had no role in study design, data collection and interpretation, or the decision to submit the work for publication.

Ethics

Animal experimentation: This study was performed in strict accordance with the recommendations in the Guide for the Care and Use of Laboratory Animals of the National Institutes of Health. All of the animals were handled according to approved institutional animal care and use committee protocols (#18-02) of the University of Oregon. All surgeries were performed under isoflurane or urethane anesthesia and every effort was made to minimize pain, distress, or discomfort.

Senior Editor

  1. Ronald L Calabrese, Emory University, United States

Reviewing Editor

  1. Jan-Marino Ramirez, Seattle Children's Research Institute and University of Washington, United States

Reviewer

  1. Jan-Marino Ramirez, Seattle Children's Research Institute and University of Washington, United States

Publication history

  1. Received: January 22, 2019
  2. Accepted: March 21, 2019
  3. Accepted Manuscript published: March 22, 2019 (version 1)
  4. Version of Record published: April 15, 2019 (version 2)

Copyright

© 2019, Hocker et al.

This article is distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use and redistribution provided that the original author and source are credited.

Metrics

  • 332
    Page views
  • 48
    Downloads
  • 0
    Citations

Article citation count generated by polling the highest count across the following sources: Crossref, PubMed Central, Scopus.

Download links

A two-part list of links to download the article, or parts of the article, in various formats.

Downloads (link to download the article as PDF)

Download citations (links to download the citations from this article in formats compatible with various reference manager tools)

Open citations (links to open the citations from this article in various online reference manager services)