Introduction

Premature infants have unstable breathing characterized by apnea of prematurity (AOP) and periodic breathing.1 The ventilatory responses to changes in oxygen and CO2 tension are mainly mediated by peripheral and central chemoreceptors, respectively. In premature infants, in comparison to term infants, the central response to an increase in CO2 is attenuated, while the contribution of peripheral chemoreceptors on baseline breathing is accentuated; both contribute to breathing instability.2 The sensitivity of central chemoreceptors improves with maturation, leading to an increase in breathing stability.3 Moreover, CO2 is a potent respiratory stimulant, and when administered at low doses to premature infants, it stabilizes breathing and decreases the frequency of apnea.4 However, in extremely low birth weight infants, breathing is more stable during the first week of age, evidenced by low frequency of intermittent hypoxic episodes and periodic breathing. Breathing then becomes more unstable between the 2nd and 4th weeks of postnatal life evidenced by increased frequency of intermittent hypoxic episodes and periodic breathing.5

Biological factors produced by the placenta, such as prostaglandins, are known to inhibit fetal chest wall movement and depress breathing postnatally.6 Leptin is another biological factor that is produced by the placenta and adipose tissue.7 While leptin is best known for its central regulatory effect on satiety and energy8 by binding to receptors in the hypothalamus,9 it is also a potent respiratory stimulant, increasing CO2 sensitivity in both humans and preclinical models.10 Adults who are leptin resistant or individuals with anorexia who have endogenously low leptin levels hypoventilate.11,12

At birth, leptin levels fall precipitously.13 Premature infants at the greatest risk for unstable breathing are the smallest and youngest, with the lowest plasma levels of leptin.2 The role of leptin on CO2 chemosensitivity during development is unknown. Neurons within the central respiratory network, specifically within the Nucleus Tractus Solitarius (NTS) and locus coeruleus (LC) are CO2 responsive,14 and express leptin receptors.15 We hypothesized that the emergence of unstable breathing after the first several days of life in premature infants might in part be mediated by the drop in leptin levels decreasing CO2 chemosensitivity. We tested this hypothesis in freely moving newborn rats, a well described model of respiratory control during development, by determining the effect of exogenous leptin on CO2 sensitivity in animals within the first 3 weeks of postnatal development. Moreover, we determined whether exogenous leptin increased the expression of phosphorylated Signal Transducer and Activator of Transcription (pSTAT) and suppressor-of-cytokine signaling (SOCS3) pathways, two well described signaling pathways activated by leptin16,17 in the LC and NTS.

Materials and methods

Animals

Time-dated pregnant Sprague Dawley rats from Charles River Laboratories (St-Constant, Quebec, Canada) were used. The animals were housed at St. Michael’s Research Vivarium and kept in cages in standard laboratory conditions (circadian day and night light of 12:12 h, 22 °C and 40% humidity) with free access to standard chow and water. The experimental protocols were approved by the Animal Care Committee at St. Michael’s Hospital that comply with the ARRIVE guidelines and by the Animal Care Committee of the Hospital for Sick Children, Toronto, Ontario, Canada, and were performed in accordance to the National Institutes of Health guide for the care and use of laboratory animals (NIH Publications No. 8023, revised 1978) and with the Canadian Council on Animal Care guidelines

Experiment protocols

To determine the effect of exogenous leptin on ventilatory responses to hyperoxia and hypercapnia

Pups were kept with the nursing dam and littermates until the day of the experiment. A total of 43 rats with both sexes represented equally in each treatment group were studied at p4 (range 3–5, n = 24), and p21 (range 19–22, n = 19). On the day of the experiment, pups were injected intraperitoneal (IP) with either normal saline (SAL), leptin 6 µg/g (Lep6) or 9 µg/g (Lep9) (murine Leptin, Peprotech, Quebec, Canada); n = 6–8 animals at each age for each treatment group. Thereafter, pups were returned to their cage for 30 min and then placed unrestrained in either a 500 mL (p4) or 1000 mL (p21) plethysmograph chamber for 30 min. Pups were sequentially exposed to room air (21% FiO2), hyperoxia (40% FiO2), hyperoxic hypercapnia (40% FiO2/5% FiCO2) and then returned to hyperoxia (40% FiO2) for a total of 15 min per gas exposure. To eliminate the contribution of the peripheral arterial chemoreceptors to the hypercapnic response, animals were exposed to 40% FiO2 followed by 40% FiO2/5% FiCO2). Animals remained in the plethysmograph for approximately 90 min. For additional details, please see Supplementary material.

Respiratory measurements

Respiratory measurements were obtained in freely moving, unrestrained animals using whole-body plethysmograph system (Buxco, DSI), as previously described.18 Respiratory rate (FR, breaths per minute), inspiratory time (TI), expiratory time (TE), and tidal volume (VT, fixed units/100 g) were derived from pressure changes using LabChart Pro (v8.1.7, ADInstruments), by peak-to-peak analysis. The Hyperoxic Ventilatory Response was calculated as the change in the ventilatory parameters from baseline (room air) to hyperoxia (40% FiO2) and is expressed by the percentage change in the respiratory parameters. The Hypercapnic Ventilatory Response (HCVR) was determined by the change in the ventilatory parameters from hyperoxia to hyperoxic-hypercapnia and is represented as the percentage change. Apnea was defined as cessation of breathing for more than two breathing cycles and was measured by counting recorded pauses in breathing that were longer than 1 s, over 5 min of stable breathing during each gas exposure.19 Apnea index was defined as the number of apnea events per hour.

To determine the effect of exogenous leptin on pSTAT and SOCS3 expression in brainstem nuclei

A total of 33 rat pups of both sexes at p4 and p21 were used. Each pup was injected IP with either saline (p4 n = 6, p21 n = 5), Lep6 (p4 n = 6, p21 n = 5) or Lep9 (p4 n = 6, p21 n = 5). One hour later, pups were euthanized by euthanyl IP injection. Blood was collected by direct cardiac puncture, followed by decapitation and removal of the brain for microdissection of the hypothalamus, LC and NTS.20 Blood was separated and frozen and later assayed for leptin levels using ELISA. Brain tissue was flash frozen using liquid nitrogen and kept in −80 °C until later processing for protein analysis by western blot.

ELISA

Leptin levels were detected using rat leptin ELISA kit (Invitrogen, ThermoFisher, Canada). For additional details, please see Supplementary material.

Western Blot

Total protein for p-STAT3 and SOCS-3 normalized β-actin expression was performed by using standing western blotting techniques previously described21 and detailed in the Supplementary materials. Monoclonal mouse antibodies (Santa Cruz Biotechnology, Canada): STAT3 (sc-8019), p-STAT3 (sc-8059), and SOCS3 (sc-73045) at 1:500 dilutions and secondary antibody (Anti-mouse IgG, HRP-linked Antibody, New England BioLabs Cell Signaling #7076 S) and enhanced chemiluminescence was used for band detection as previously described were used.22

Statistical analysis

To detect association between numerical variables, data were analyzed using two-way ANOVA with condition as repeated variable, followed by Holm-Sidak t-test for individual comparisons of the means (SigmaPlot for Windows Version 11.0, 2008 System Software, Inc.). For non-normal distributed variables, we used the nonparametric Mann–Whitney U t-test. To detect an association between continuous variables, we used linear regression analysis. Statistical significance was defined as P < 0.05. Data presented at mean ± SE or median and IQR when appropriate. The coefficient of variance (CV) was calculated for VE using descriptive statistics. The study was not powered to detect sex difference.

Results

Mean body weight and core body temperature were similar in rats treated with saline (SAL), leptin 6 µg/g (Lep6) and leptin 9 µg/g (Lep9) at both p4 and p21. Since animals treated with Lep9 had a large variance in responses, we have included all responses from the 3 treatment groups in the tables, but only the responses from animals treated with saline and Lep6 in the graphs. Fig. 1 is a representative tracing of ventilatory responses from one animal at each age group, treated with SAL and Lep6.

Fig. 1: Representative ventilatory tracing of rats using whole-body plethysmography in postnatal days 4 and 21.
figure 1

Animals were treated with saline or leptin (6 µg/g). Scale: small square: 0.2 s, big square: 1 s. P postnatal day.

Full size image

Leptin eliminates maturational effects of hyperoxic induced decline in F R

Transition from room air to hyperoxia induced respiratory depression in p4 and p21 SAL-treated animals and p21 Lep-treated animals by decreasing FR while VE did not change (Table 1). Leptin 6 µg/g decreased the percent reduction in the hyperoxic FR response in p4 animals when compared to the response in SAL treated animals (Lep6 7.5 ± 4% vs SAL 24 ± 4% P = 0.04, Fig. 2). With maturation, the percent reduction in hyperoxia-induced respiratory depression decreased in p4 to p21 in saline treated animals. However, leptin treatment in p4 animals eliminated this maturational difference such that p21 and p4 animals had the same reduction in hyperoxic decline in FR (7.5 ± 4% at p4 vs 11 ± 4% at p21, P = 0.9). In animals at p21, leptin had no effect on the hyperoxia-induced decline in frequency. The decrease in FR with the transition from room air to hyperoxia was associated with an increase in apnea index (AI) in all age groups except in p21 animals treated with Lep9 (Table 2). Leptin treatment did not modify the magnitude of the changes in AI from room air to hyperoxia in any of the groups. Finally, breathing variability, as expressed by FR coefficient variation (FRCV) in room air was higher in p4 (50 ± 5) than in animals at p21 (25 ± 2; P < 0.05 Fig. 3); exogenous leptin did not modify this difference in FR CV in either age group.

Table 1 Ventilation parameters (tidal volume, breathing frequency and minute ventilation) of rats at postnatal days 4 and 21 exposed to room air and hyperoxia (40% FiO2).
Full size table
Fig. 2: Effect of leptin on the hyperoxic ventilatory response of rats at postnatal days 4 and 21.
figure 2

The ventilatory response is expressed as percentage change in breathing frequency of animals transitioned from room air to hyperoxia (40% FiO2). Animals were treated with saline (p4: n = 7, p21: n = 5, white bars) or leptin (6 µg/g, p4: n = 8, p21: n = 6, black bars). Saline treated animals at p4 had greater percent reduction in breathing frequency in response to hyperoxia compared to animals at p21 (*P = 0.01) and p4 animals that were treated with leptin (**P = 0.04). Leptin treatment had no effect on hyperoxic –induced reduction in breathing frequency at p21 but did at p4. Data are presented as mean ± SEM (2-way ANOVA). P postnatal age, FR breathing frequency.

Full size image
Table 2 Apnea Index at (a) room air and hyperoxia (40% FiO2) and at (b) hyperoxia (40% FiO2) and hypercapnia (40% FiO2/ 5% FiCO2) of rats at postnatal days 4 and 21.
Full size table
Fig. 3: Effect of leptin on respiratory variability.
figure 3

Coefficient of variation (CV) of breathing frequency was calculated in room air for rats at postnatal days 4 and 21. In saline treated animals, CV of breathing frequency FR in p4 animals was higher when compared to animals at p21 (*P < 0.05 p4 vs p21). Leptin had no effect on CV of FR at either age group. Data are presented as mean ± SEM (2-way ANOVA). Saline (p4: n = 7, p21: n = 5, white bars); leptin (6 µg/g, p4: n = 8, p21: n = 6, black bars). P postnatal age, FR breathing frequency.

Full size image

Leptin increases Hypercapnic Ventilatory Response (HCVR) in p21 animals

When transitioned to hyperoxia/hypercapnia (40% O2/5% CO2) from hyperoxia (40%), VE and VT increased in animals at all age and treatment groups and FR increased in all age and treatment groups except p4 Lep6 animals (Table 3). Exogenous leptin only modified the HCVR in p21 animals. Specifically, in leptin-treated animals at p21, FR increased by 50 ± 4% compared to 30 ± 4% in saline-treated animals (P < 0.001, Fig. 4a). The apnea index did not differ between control or leptin exposed animals during hyperoxia and hyperoxia/hypercapnia exposure (Table 2b).

Table 3 Ventilation parameters (tidal volume, frequency, and minute ventilation) of rats at postnatal days 4 and 21 exposed to hyperoxia (40% FiO2) and hyperoxia/ hypercapnia (40% FiO2/ 5% FiCO2).
Full size table
Fig. 4: Leptin increases the hypercapnic ventilatory response (HCVR) at postnatal day 21, but not at postnatal day 4.
figure 4

The ventilatory response is expressed as percentage increase in (a) breathing frequency (FR) and (b) minute ventilation (VE) in animals transitioned from hyperoxia (40% FiO2) to hyperoxia-hypercapnia (40% FiO2 /5% FiCO2). In saline treated animals the percent change in HVCR for FR and VE did not differ between p21 and p4. In leptin treated animals, the percent change in FR and VE was greater in p21 vs p4 animals (**P < 0.001). Leptin increased the percent change in FR of HVCR in p21 but not in p4 animals (*P = 0.004). Data are presented as mean ± SEM (2-way ANOVA). Saline (p4: n = 8, p21: n = 6) or leptin (6ug/g, p4: n = 8, p21: n = 7). P postnatal age.

Full size image

Exogenous leptin increases plasma leptin levels in p4 animals

We determined the effect of exogenous leptin on plasma leptin levels with ELISA at p4 and p21, (n = 4–6, each group). In animals at p4, at 1 h after IP injection of leptin (6 µg/g or 9 µg/g; mean ± SE) plasma leptin levels were 90 ± 1.3 ng/mL and 90 ± 1.4 ng/mL, respectively. These levels are much, much higher than endogenous leptin levels of 1.3 ± 0.3 ng/mL in p4 animals given SAL (SAL vs. Lep6 and Lep9, P < 0.001). However, in animals at p21, exogenous leptin (6 µg/g or 9 µg/g) did not significantly change endogenous leptin levels (2.0 ± 0.05 ng/mL). We suspect that inability to detect changes in plasma leptin levels in the animals at p21 after IP exogenous leptin administration could be from leptin binding to leptin receptors on adipocytes within the omentum.

Leptin effect on signal transduction in the hypothalamus, NTS and LC

Lastly, we determined the effect of exogenous leptin (6 µg/g) on JAK-STAT3 signaling pathway in the hypothalamus, NTS and LC from p4 and p21 animals, by measuring protein levels of phosphorylated STAT (pSTAT) and SOCS3.

Exogenous leptin effect on pSTAT expression

In SAL-treated animals, p-STAT protein levels were higher in the hypothalamus of p21 compared to p4 animals (P < 0.05, Fig. 5a), while pSTAT expression in the NTS and LC (Fig. 5b, c) did not differ between the two groups. In comparison to SAL-treated animals at p21, exogenous leptin increased pSTAT expression in the hypothalamus, NTS, and LC (Fig. 5a–c) with the greatest fold increase in protein expression observed in the hypothalamus [(1.63 ± 0.11 in SAL vs. 2.59 ± 0.12 in Lep6 treated animals at p21) P < 0.001, Fig. 5a]. However, in p4 animals, exogenous leptin augmented pSTAT expression only in the LC, with 1.33 times greater expression in comparison to SAL-treated animals (P = 0.029, Fig. 5c).

Fig. 5: Effect of leptin on pSTAT3 and SOCS3 protein expression in the brain.
figure 5

Fold change in protein expression of pSTAT and SOCS3 in leptin treated animals compared to control. P4 saline-treated animals represent baseline values. Hypothalamic (a, d), NTS (b, e) and LC (c, f) tissues were isolated from rats at P4 and 21 who were treated with saline (SAL, p4: n = 6, p21: n = 5) or leptin (6 µg/g, p4: n = 6, p21: n = 5). Western blot was used to determine the level of protein expression for pSTAT and SOCS3 normalized to β-actin. Data are presented as mean ± SEM (2-way ANOVA). Protein bands from each of the animals are depicted beneath the graph. Closed circles and triangles represents animals treated with saline or leptin, respectively. pSTAT expression: pSTAT expression in SAL treated rats was higher at p21 compared to p4 rats only in the hypothalamus (**P < 0.05, a). Leptin induced pSTAT expression in p4 rats only in the LC (P = 0.029, c), while in p21 rats leptin increased the expression of pSTAT in the hypothalamus (*P < 0.001, a), NTS (*P < 0.001, b) and LC (*P < 0.001, c). SOCS3 expression: At p21 compared to p4, SOCS3 expression in SAL treated rats was higher only in the hypothalamus (**P < 0.05, d). However, in the LC, p21 rats had lower SOCS3 expression compared to p4 rats (**P < 0.05). Leptin induced SOCS3 expression in p4 rats in the NTS (**P < 0.05, e) and LC (*P < 0.001, f), while in p21 rats it increased the expression in the hypothalamus (*P < 0.001, d), NTS (*P < 0.001, e) and LC (**P < 0.05, f). P postnatal age, NTS nucleus tractus solitarii, LC locus coeruleus.

Full size image

Exogenous leptin effect on SOCS3 expression

In SAL-treated animals, SOCS3 protein expression was higher in the hypothalamus of p21 compared to p4 animals (P < 0.05, Fig. 5d), while SOCS3 expression in the NTS (Fig. 5e) did not differ between the two groups. In contrast, SOCS3 expression in the LC of p21 animals was lower compared to p4 animals (P < 0.05, Fig. 5f). Similar to its effect on p-STAT expression, in p21 animals, exogenous leptin increased SOCS3 protein levels in the hypothalamus NTS and LC (Fig. 5d–f) with the greatest fold change in SOCS3 expression observed in the hypothalamus [(1.95 ± 0.34 at SAL-p21 vs. 4.46 ± 0.34 at Lep6-p21) P < 0.001, Fig. 5d]. However, in p4 animals, exogenous leptin augmented SOCS3 expression in the NTS and LC but not in the hypothalamus (Fig. 5e, f). If increased pSTAT expression signifies activation and SOCS3 signifies inhibition of the pSTAT pathway, then the summation of the signaling within the NTS and LC suggest that these neurons may not be depolarizing in p4 animals vs. p21 animals—in the presence of leptin, signaling summation favors activation.

Discussion

In addition to regulating appetite and metabolism in the hypothalamus,9 leptin also acts as a respiratory stimulant, modulating CO2 chemosensitivity11,23 via activation of the pSTAT and SOCS3 pathways16,17 in the NTS and LC.23,24 Premature infants have decreased ventilatory responses to CO2 and an increased frequency of apnea.1 Leptin levels acutely fall after delivery and apnea frequency increases during the first several weeks of postnatal life.2 The effect of leptin on CO2 chemosensitivity and signal transduction in the central respiratory network during development has not been described. Here we show that exogenous leptin augments the hypercapnic ventilatory response and increases pSTAT and SOCS3 protein expression in the LC, NTS, and hypothalamus in Sprague Dawley rats at p21. In contrast, exogenous leptin did not alter hypercapnic ventilatory response in animals at p4, but it did increase protein expression of pSTAT in the LC, and SOCS3 in both the LC and NTS but not in the hypothalamus. Lastly, exogenous leptin reduced chemosensitivity to hyperoxia only in the younger animals with no significant effect on hyperoxia-induced apnea or variability in respiratory frequency in either age group. Our data suggest that leptin modulates CO2 responsiveness in newborn rodents only after 3 weeks of postnatal age, similar to when leptin signaling in the hypothalamus modulates satiety.

Peripheral and central chemoreceptors’ contribution on breathing changes with development.25 Peripheral arterial chemoreceptors in the carotid body are exquisitely sensitive to changes in oxygen tension. Hypoxic chemosensitivity reaches adult levels by 3 weeks of postnatal age in infants26,27 and rats.28 However, peripheral arterial chemoreceptors’ effect on baseline breathing decreases with maturation, supported by the observation that periodic breathing decreases with maturation;29 the youngest infants have the greatest reduction in ventilation in response to hyperoxia.30 Thus, carotid body influence on baseline breathing is ∼20% in premature infants vs. 5% in adults.29 Furthermore, animals at p4 had greater hyperoxic reduction in breathing frequency than p21 animals, and the CV of breathing frequency in room air was 2-fold greater in animals at p4 than at p21. These findings support a greater influence of peripheral arterial chemoreceptors on baseline breathing at p4 animals than at p21. Exogenous leptin decreased hyperoxia-induced ventilatory decline at p4 but not p21 and the percent change in the reduction in the FR in p4 and not p21 animals.

The long isoform of the leptin receptor is present on carotid body type 1 cells.31 Exogenous leptin increases the ventilatory responses to changes in oxygen tension, which are abolished by carotid sinus nerve denervation. Similar conclusions were made from experiments transfecting LpR into the carotid body of LepR deficient mice.32 Since our focus centered on leptin’s effect on CO2 chemosensitivity, we did not determine whether leptin modified hypoxic ventilatory responses in either age group, but we did observe that exogenous leptin blocked hyperoxia-induced depression of FR at p4. These findings, given that leptin did not affect hypercapnic ventilatory responses in p4 animals (discussed below), suggest that leptin decreases the contribution of the carotid body on baseline breathing in p4 animals. Moreover, we suggest that leptin in p4 animals decreases overall responsiveness of the ventilatory system, most likely by altering the sensitivity of peripheral arterial chemoreceptors versus the sensitivity of central chemoreceptors.33 However, our results do not rule out that exogenous leptin may also affect central CO2 chemosensitivity.

Maturation of the central chemoreceptor response to pCO2 and pH, similar to that of peripheral arterial chemoreceptor, also occurs during early postnatal development. Putman et al. showed that CO2 chemosensitivity in rats is high at birth, reaches a nadir by the end of the first week, gradually increases p8, and reaches adult responsiveness by the third postnatal week.34 We show that the hypercapnic ventilatory response was similar in animals at p4 and p21. While the leptin surge occurs in rodents around days 8 (rats) and 10 (mice), we found that CO2 responsiveness in animals at p7–9 was highly variable with or without exogenous leptin. Only in animals at p21 did exogenous leptin consistently augment ventilation (via frequency; no effect on tidal volume). Even after increasing plasma leptin levels by 45-fold, the hypercapnic ventilatory response was not altered in p4 animals. We conclude that the hypercapnic ventilatory response in newborn rats in the first postnatal week is not modulated by leptin. At this age, leptin’s primary role is to coordinate the development of neuronal circuits and peripheral organs.35,36

With maturation and increase in fat mass, plasma leptin levels rise in infants and children.37,38 Leptin’s primary role in children and adults is to regulate food intake and energy expenditure, binding to leptin receptors on neurons in the hypothalamus. However, during early postnatal development in rodents and piglets, leptin also regulates organ development and branching morphogenesis.37,39,40 In rodents, through p21, leptin signaling provides trophic effects on the development of neurocircuitry that will regulate feeding behavior.36 The leptin anorexic effect in the hypothalamus develops only after weaning in rodents,41,42 despite the leptin surge36 and upregulation of leptin receptors in the micro vessels of the blood brain barrier.43 This anorexic effect after weaning corresponds with the leptin-augmented respiratory response that we observed at p21 but not at p4.

Our findings support exogenous leptin’s effect on hypercapnic respiratory response. In the hypothalamus, this is mediated by activation of the excitatory JAK-STAT pathway and the inhibitory SOCS3 pathway.44 Leptin stimulates breathing by binding to LepR in the dorsomedial hypothalamus.45 Corroborating the physiological responses, exogenous leptin increased pSTAT and SOCS3 expression in the hypothalamus of animals at p21 but not p4 animals. Similarly, Cottrell and colleagues found that leptin induced SOCS3 expression only at p14 and not p4 in the ventromedial hypothalamic nucleus (VMH). These differences coincided with changes in LepR expression in the VMH.36 The emergence of the anorexic response could also be partially explained by changes in leptin binding activity. Leptin binding to LepR soluble receptor in rats increases from fetal life to 4 weeks postnatally, making less leptin available to bind to the long isoform of the LepR.46 LepR expression increases in the arcuate nucleus of the rat hypothalamus during the first 100 days of life.47

Neurons in the NTS and the LC contribute to hypercapnic ventilatory response.14 Neurons in the NTS express LepR48 and leptin microinjection into the NTS augments respiratory response to hypercapnia.49 We show here the effect of exogenous leptin on protein expression of pSTAT and SOCS3 in the NTS during the first 3 postnatal weeks. Exogenous leptin increased pSTAT and SOCS3 expression in the older animals. In younger animals, leptin only increased SOCS3 expression at p4 with no effect on pSTAT expression in the NTS, suggesting leptin inhibits neurons within the NTS and LC. Although leptin upregulation of pSTAT and SOC3 has been correlated with adult brain neuronal activity,9 it is unclear whether the same pattern of expression mediates excitation or inhibition in cells and neurons in the NTS and LC. We observed a predominance of an inhibitory signature in the LC of p4 animals corresponding to the absence of a hypercapnic response to exogenous leptin. The increased pSTAT and SOCS3 expression in the LC at p21 could be due to enhanced activity of LepR-expressing neurons projecting from the hypothalamus to the LC, which, at p21, are more responsive to exogenous leptin than at p4.50

To our knowledge, this is the first study in a newborn model of respiratory development to test the effect of leptin as a respiratory stimulant augmenting CO2 chemosensitivity. We demonstrate the effect of leptin is delayed until 3 weeks of age, like the known anorexic effect of leptin in the hypothalamus at this age. Our physiologic findings are supported by the changes in expression of pSTAT and SOCS3 in the central respiratory network. Our study has several limitations: (1) we did not do single cell patch clamping in the different cells within the hypothalamus, NTS, or LC; (2) we did not measure cerebrospinal fluid levels of leptin or give it intranasally, which may have been more effective in the younger animals, especially if there is a higher expression of soluble leptin receptor at this age; and, (3) we elected to express changes in protein expression of pSTAT normalized to β-actin and not total STAT, per the only other study to quantitate change in early development hypothalamic pSTAT expression.44 Nevertheless, we are certain that exogenous leptin does not alter hypercapnic chemosensitivity in newborn rats during the first 4 postnatal days, but it does appear to eliminate the respiratory depression associated with hyperoxic exposure at this age.

If our findings can be extrapolated to human infants, elevated leptin levels at birth in premature infants are unlikely to contribute to hypercapnic ventilatory drive. However, exogenous leptin decreased carotid body contribution to baseline breathing and the variability of respiratory frequency at baseline at p4. Thus, the natural decline in leptin and appearance of more unstable breathing during the 1st and 2nd week of life after birth might be mediated by removal of the placenta, the main source of plasma leptin in premature infants. The youngest and smallest infants can have little-to-no leptin until fat mass increases at around 34 weeks’ postnatal age.2,38 The consequences of low leptin levels during critical periods of organ development in premature infants has been understudied but is likely of substantial significance.