Authors: Gabriel C. Ripamonte, Elisa M. Fonseca, Alana T. Frias, Luis Gustavo A. Patrone, Heloísa H. Vilela-Costa, Kaoma S. C. Silva, Raphael E. Szawka, Kênia C. Bícego, Hélio Zangrossi, Jr, Nicholas W. Plummer, Patricia Jensen, Luciane H. Gargaglioni
Categories: Article, hypercapnia, escape, catecholamines, suffocation alarm, breathing
Source: Progress in neuro-psychopharmacology & biological psychiatry
Authors: Gabriel C. Ripamonte, Elisa M. Fonseca, Alana T. Frias, Luis Gustavo A. Patrone, Heloísa H. Vilela-Costa, Kaoma S. C. Silva, Raphael E. Szawka, Kênia C. Bícego, Hélio Zangrossi, Nicholas W. Plummer, Patricia Jensen, Luciane H. Gargaglioni
CO2 exposure has been used to investigate the panicogenic response in patients with panic disorder. These patients are more sensitive to CO2, and more likely to experience the “false suffocation alarm” which triggers panic attacks. Imbalances in locus coeruleus noradrenergic (LC-NA) neurotransmission are responsible for psychiatric disorders, including panic disorder. These neurons are sensitive to changes in CO2/pH. Therefore, we investigated if LC-NA neurons are differentially activated after severe hypercapnia in mice. Further, we evaluated the participation of LC-NA neurons in ventilatory and panic-like escape responses induced by 20% CO2 in male and female wild type mice and two mouse models of altered LC-NA synthesis. Hypercapnia activates the LC-NA neurons, with males presenting a heightened level of activation. Mutant males lacking or with reduced LC-NA synthesis showed hypoventilation, while animals lacking LC noradrenaline present an increased metabolic rate compared to wild type in normocapnia. When exposed to CO2, males lacking LC noradrenaline showed a lower respiratory frequency compared to control animals. On the other hand, females lacking LC noradrenaline presented a higher tidal volume. Nevertheless, no change in ventilation was observed in either sex. CO2 evoked an active escape response. Mice lacking LC noradrenaline had a blunted jumping response and an increased freezing duration compared to the other groups. They also presented fewer racing episodes compared to wild type animals, but not different from mice with reduced LC noradrenaline. These findings suggest that LC-NA has an important role in ventilatory and panic-like escape responses elicited by CO2 exposure in mice.
Panic disorder (PD) is a psychiatric disorder that affects about 2% of the world population and is defined by the emergence of sudden panic attacks, which are characterized by intense stress accompanied by cardiorespiratory symptoms such as shortness of breath and tachycardia (Nardi et al., 2013; De Jonge, 2016; Spiacci et al., 2018; Okuro et al., 2020). In fact, an extensive collection of evidence suggests a likely connection between PD and respiratory distress, including respiratory challenges that are largely used to trigger panic attacks, showing a relation between emotional states and respiration (Briggs et al., 1993; Ziemann et al., 2009; D’amato et al., 2011; Nardi et al., 2013; Taugher et al., 2014; Leibold et al., 2016; Vollmer et al., 2016). Stress-related respiratory disorders cover a broad range of clinical symptoms including hyperventilation, panic attacks and even asthma. Despite their clinical importance, the origins and pathophysiology of stress-related respiratory disorders remain poorly understood (Klein, 1993; Nardi et al., 2013; Kinkead et al., 2014).
Changes in arterial O2 and/or CO2 levels promote ventilatory adjustments due to activation of central and peripheral chemoreceptors to return arterial blood gases close to their set point (Feldman et al., 2013). Nevertheless, a sudden drop in O2 and/or increase in CO2 can also generate significant neuroendocrine and emotional/behavioral reactions that can be very strong (Kinkead et al., 2014). Suffocation is perhaps the most powerful stressor and whenever an escape response is possible it may be the best defense strategy to avoid an improper respiratory environment. Indeed, clinicians established a link between respiration and anxiety, and nowadays, key theories of the psychopathology of anxiety (including PD) focus on respiratory control and its related monitoring system (Klein, 1993; Abelson et al., 2010). Therefore, several studies have sought to analyze the responses of PD patients to the inhalation of a gaseous mixture rich in CO2 (Nardi et al., 2013; Wiese and Boutros, 2019; Okuro et al., 2020). These elevated levels of CO2 correspond precisely to the range in which we would expect to see the activation of a suffocation alarm (Preter and Klein, 2008) and, consequently, the induction of anxiety/panic disorders (Griez et al., 2007). In fact, the exposure of mice to 20% CO2 evokes defensive behaviors that are suggestive that the animals are indeed experiencing a highly aversive situation that resembles the triggering of a panic attack (Spiacci et al., 2018).
Stressful experiences incite many brain regions to produce a coordinated physical and psychological response to a challenge (Blanchard et al.,1993). Locus coeruleus (LC) noradrenergic neurons project widely to these brain areas (Schwarz et al., 2015; Robertson et al., 2013), and they are also linked to both physical and emotional responses to stress (Berridge and Waterhouse, 2003; Valentino and Van Bockstaele, 2008) and aversive memory consolidation (Roozendaal et al., 2008). In reality, LC activation during a stressful situation induces an anxiety state response that allows the individual to maintain high attention, facilitate sensory processing and enhance decision-making functions in order to increase memory consolidation during stressful experiences (Beeridge et al., 2003; Sara and Bouret, 2012).
Imbalance in the modulation of the LC-NA system is associated with numerous psychiatric pathologies, such as attention deficit hyperactivity disorder, neurodegenerative diseases, post-traumatic stress disorder, depression and anxiety, including PD (Bangasser and Valentino, 2014; Fortress et al., 2015; Isingrini et al., 2016; Weinshenker, 2018). In clinical studies, anomalous noradrenergic regulation was reported in PD patients and during panic attacks (Charney, et al., 1990; Bremner et al., 1996; Balaban, 2002; Dierssen et al., 2002). Elevated noradrenergic activity enhances anxiety-like behavioral responses and inappropriate activation of the LC, which may participate in the exaggerated stimulus-responsiveness and increased emotionality seen in patients with stress or anxiety disorders (Priolo et al., 1991; Aston-Jones et al., 1994; Goddard and Charney, 1997). On the other hand, antidepressant treatment which is effective in panic disorder patients decreases LC firing and tyrosine hydroxylase expression (West et al., 2009). Also, behavioral manipulations that decrease stress, such as postnatal handling modulate the responses of the noradrenergic system (Escorihuela et al., 1995; Escorihuela et al., 1995; Baamonde et al., 2002).
Locus coeruleus neurons function directly as respiratory CO2/pH chemosensors (Coates et al. 1993; Gargaglioni et al., 2010). The LC neurons are of particular interest in CO2 challenges since >80% of these neurons are found to be chemosensitive, responding to hypercapnia with an increased firing rate (Pineda and Aghajanian, 1997; Oyamada et al., 1998; Filosa et al., 2002). Indeed, a reduction of approximately 80% of the noradrenergic neurons of LC was associated with an approximately 64% decrease in ventilation in response to 7% CO2, indicating once again that the LC exerts a profound CO2-induced drive to breathe (Biancardi et al., 2008). However, there is no study showing the participation of LC-NA neurons in ventilatory and behavioral responses evoked by higher levels of CO2 (20%), a stimulus that evokes a typical repertoire of behaviors interpreted as active panic-like escape response (Spiacci et al., 2018).
In the current study, our primary objective was to examine the activation pattern of LC-NA neurons, as indicated by c-Fos expression, in both male and female mice when exposed to 20% CO2. c-Fos has been established as a reliable marker for assessing neuronal activation and is, therefore, a valuable tool for mapping the functional aspects of anxiety-related neurocircuitry. Additionally, we sought to assess the involvement of noradrenergic neurons located in the LC in relation to behavioral, respiratory and metabolic parameters during a CO2-induced panic attack, utilizing mice with conditional loss of LC noradrenaline.
C57BL/6J male and female mice (10–12 weeks old) were used for immunohistochemical analysis of LC-NA neuronal activation. Experimental mice for other experiments were generated by crossing a conditional knockout allele of dopamine β-hydroxylase (Dbh^flx^; Wilson et al., 2023) to En1^Cre^ (Kimmel et al., 2000). Prior to intercrossing, Dbh^flx^ mice had been crossed to C57BL/6J for ≥7 generations, and En1^Cre^ mice for >20 generations. En1^Cre/wt^ Dbh^flx/flx^ (Dbh^LC-null^) mice lack Dbh expression required for noradrenaline synthesis in the LC and have reduced Dbh expression in other noradrenergic neurons, while En1^wt/wt^ Dbh^flx/flx^ (Dbh^hypo^) mice have reduced Dbh expression in LC and other noradrenergic neurons (Wilson et al., 2023). En1^wt/wt^ Dbh^wt/wt^ mice (hereafter designated Dbh^wt/wt^) served as wild-type controls. All experimental cohorts included both sexes. The animals were submitted to a 12/12 h light/dark cycle, conditioned at a controlled temperature of 25 ± 1 °C and had food and water ad libitum. The experimental protocols involved the random selection of animals, ensuring that each genotype was represented in experiments conducted on the same day. All experiments were conducted between 8 AM and 6 PM at the Department of Animal Morphology and Physiology of the Faculty of Agricultural and Veterinary Sciences of Jaboticabal. All protocols received approval from the local ethics committee (CEUA – protocol n° 3340/20) and adhere to the guidelines established by the National Council for the Control of Animal Experimentation (CONCEA. The experiments are in compliance with the ARRIVE guidelines and were be carried out in accordance with the U.K. Animals (Scientific Procedures) Act, 1986 and associated guidelines, EU Directive 2010/63/EU for animal experiments, or the National Research Council’s Guide for the Care and Use of Laboratory.
To assess neuronal activation, C57BL/6J mice underwent a one-hour daily habituation period inside the experimental chamber ventilated with atmospheric air for three consecutive days. This consistent timing before the experiment was crucial in preventing any nonspecific c-Fos expression. On the fourth day (after the 3 days of habituation), the animals were exposed to normocapnic normoxia (atmospheric air) or hypercapnic normoxia (20% CO2 for 15 min). At the end of the experiment, each animal was kept in the chamber ventilated with room air for another 60 minutes for c-Fos protein expression. After this period, the animal was deeply anesthetized with an overdose of isoflurane and perfused by the left cardiac ventricle with phosphate buffer solution (PBS - 0.01 M; pH 7.4), followed by 4% paraformaldehyde (PFA - diluted in PB phosphate buffered solution - 0.1 M; pH 7.4). The brain was removed from the skull and stored in 4% PFA at 4°C for 12 hours and then kept in a 30% sucrose solution dissolved in PBS (0.01 M; pH 7.4) at 4°C for at least 24 hours. The brain was then immersed in isopentane cooled in carbonic ice for freezing, and embedded in Tissue Tek O.C.T.; then, serial sections (40 μm) of the LC were made in a cryostat (Leica CM 1860 - Ag Protect, Germany). To verify the activity of noradrenergic neurons in the LC, a double-labeling immunohistochemistry for tyrosine hydroxylase and c-Fos was performed.
For immunohistochemistry experiments, sections were pre-incubated for 30 min in a target retrieval solution (Dako, Glostrup, Denmark) at 70°C for an antigenic retrieval process. The sections were then washed 3 times in PBS (5 minutes per wash), pre-incubated in a 1% hydrogen peroxide solution for 3 minutes, followed by 3 more washes in PBS, and then incubated in a solution of PBST (phosphate buffer solution – PBS; triton and goat serum) for at least 1 hour, these procedures being performed at room temperature. Subsequently, the sections were incubated for 24 hours at 4°C with constant agitation in a solution containing a primary antibody (monoclonal IgG anti-c-Fos protein produced in rabbits, 1000 concentration; Sigma) diluted in PBST, then washed 3 times in PBS after 24 hours of incubation. The sections were then incubated for 2 hours in a solution containing a secondary antibody (biotinylated goat anti-rabbit IgG, 1000 concentration, Dako Cytomation, Denmark, Europe) at room temperature with constant stirring, followed by 3 washes in PBS. The sections were then incubated in an avidin-biotinylated DH-horseradish peroxidase complex (Vector, code PK-4001) for 1 hour on a shaker at room temperature. After 3 washes in PBS, the labeling of catecholaminergic neurons was visualized with a buffer containing 0.05% 3,3’ diaminobenzidine tetrahydrochloride (DAB) and 0.004% hydrogen peroxide in distilled water for 1 minute, followed by 3 quick washes in PBS. The same procedure was repeated, but this time for tyrosine hydroxylase (TH) labeling. The primary antibody used was the anti-TH monoclonal antibody produced in mice (1:8000 ratio; Sigma) diluted in PBST. The secondary antibody, in turn, was biotinylated rabbit IgG anti-mouse IgG Fc in a concentration of 1000 (Dako Cytomation, Denmark, Europe). Finally, the sections were placed on gelatinized slides, dried, and covered with a coverslip.
A temperature-sensitive tag (Biomark HPR Plus Reader, Boise, ID, United States) was surgically implanted in the abdominal cavity of the mice through an abdominal wall incision for continuous body temperature monitoring. For this, mice were anesthetized with isoflurane (5% for induction and 1% for maintenance in pure O2, using a face mask). The entire procedure lasted about 10 min. Following surgery, the animals were housed in individually controlled cages with access to unlimited water and food where they were maintained in a climate-controlled room with regulated temperature, humidity, and lighting conditions for two weeks. The animals received antibiotics (enrofloxacin, Baytril; 2,5 mg.kg^−1^ S.C., Bayer S.A., São Paulo, SP, Brazil) and analgesic injections (Flunixin Meglumine, Banamine; 10 mg.kg^−1^). On the day of the experiment, individual body temperatures (Tbs) were recorded in real time by telemetry using a Biotherm reader (Biomark HPR Plus Reader, Boise, ID, United States) and uploaded to a computer (BioTerm software). Both sensors (colonic and abdominal) were calibrated against a certified thermometer (0.1°C precision).
The 20% CO2 air challenge was used to induce panic behavioral and ventilatory responses in mice (Spiacci et al., 2018). To promote a hypercapnic environment, a gas mixer (Gas mixer GSM-3, CWE inc., USA) was used to produce the following gas 21% O2; 20% CO2; and 59% N2. The CO2 concentration was gradually increased over a period of 7 minutes until it reached a level of 20%.
Ventilation (VE) was measured by whole-body plethysmography (Patrone et al., 2018). During VE measurements, the airflow was interrupted, and the chamber remained sealed for about 2 min. The airflow was 1.8 L/min measured by a flow meter (“Mass Flow System” - MFS, Sable Systems International, Inc, Las Vegas, USA). Signals from a differential pressure transducer (TSD 160A, Biopac Systems, Santa Barbara, CA - USA), connected to the animal chamber, were collected by an amplifier (DA 100C, Biopac Systems), passing through an analog-to-digital converter, digitized in a computer equipped with a data acquisition software (AcqKnowledge MP 100, BioPac Systems, Inc., Santa Barbara, CA, USA). Volume calibration was obtained during each experiment by injecting a known volume of air into the animal’s chamber (1 mL) using a graduated syringe. Two respiratory variables were measured, respiratory rate (fR) and tidal volume (VT), the latter being calculated using the formula of Drorbaugh and Fenn (1955):
In which PT is the pressure deflection associated with each VT, PK is the pressure deflection associated with the injection of the calibration volume (VK), TA is the air temperature in the animal chamber, PB is the barometric pressure, PC is the water vapor pressure in the animal chamber, Tb is the body temperature and PR is the vapor pressure of water at Tb.
Ventilation (VE) was calculated as the product of respiratory rate (fR) and tidal volume (VT). VE and VT are presented at ambient barometric pressure conditions, at Tb and saturated with water vapor (BTPS). PC was calculated indirectly using an appropriate table (Dejours, 1981).
The method of indirect calorimetry (oxygen consumption, VO2) and push mode configuration was used to infer metabolic rate through an open respiratory system. The airflow (1.8 L/min) in the chamber was maintained by an aquarium pump for room air exposure (normocarbic normoxia), and by a gas mixer (Gas mixer GSM-3, CWE Inc., USA) for hypercapnia (20% CO2; 21% O2; 59% N2). The expired gas was dried through a small column of drierite (W. A. Hammond Drierite Co. Ltd, Xenia, OH, USA) before going through the oxygen analyzer (ADInstruments^®^). The air was continuously sampled by the O2 analyzer allowing the determination of VO2 by the data acquisition software (Power-Lab System, ADInstruments^®^/Chart Software, version 7.3, Sydney, Australia.)
Oxygen consumption (VO2) was calculated using
In which Fl is the inflow rate of air in the chamber, F’iO2 is the inspired O2 fraction, F’eO2 is the O2 fraction expired and RQ is the value referring to the energy source used by the animal.
To analyze panic attack behavioral responses, the experiments were recorded by two cameras (Logitech, USA) strategically positioned beside and above the experimental chamber, enabling the recording of the animal and its movements. The escape response analyzed included the number of jumps, the number of runs (running) and freezing (total lack of movement except breathing). All analyzed by a trained observer.
To analyze the monoamine concentrations, the animals of all experimental groups were exposed to 20% CO2, and then euthanized. The brains of male and female mice were rapidly excised and frozen in dry ice-cold isopentane. All samples were stored at −80°C. The brainstem and forebrain were separated and homogenized in a solution containing 0.15 M perchloric acid, 0.1 mM ethylenediaminetetraacetic acid (EDTA), and 0.17 μM 3,4-dihydroxybenzylamine as an internal standard. The homogenates were centrifuged for 20 min at 12,000 G and 4°C. Protein content was determined from the pellet, and the supernatant was analyzed for NA, DA, 3,4-dihydroxyphenylacetic acid (DOPAC; the main metabolite of DA), by high-performance liquid chromatography coupled to an electrochemical detector (HPLC-ED), as previously described (Aquino et al., 2017; Silva et al., 2020). The chromatography separation was carried out using a C-18 column (250 mm × 4 mm, 5 μm; Merck, Darmstadt, Germany), preceded by a C-18 pre-column (5 μm, 4 mm × 4 mm; Merck), and kept at 40°C. The mobile phase consisted of 100 mM NaH2PO4, 10 mM NaCl, 0.1 mM EDTA, 0.38 mM sodium 1-octanesulfonic acid, and 10% methanol in ultrapure water, pH 3.5. The pump flow rate was 1.0 mL/min, and the potential in the electrochemical detector (Decade II; Antec Scientific, Zoeterwoude, Netherlands) was set to +0.40 V vs. the Ag/AgCl reference electrode. All samples from each brain area were measured in the same analysis. The intra-assay coefficient of variation was less than 5% for all measured compounds. DA levels were considered to reflect neurotransmitter stock in synaptic vesicles, whereas DOPAC levels reflected neurotransmitter release (Nilsen et al., 1986; Lookingland et al., 1987). DOPAC/DA ratios were used as a measure of neurotransmitter turnover.
Each mouse was placed in the experimental apparatus (a cylindrical chamber, 18 cm in diameter x 18 cm in height) for 60 minutes for habituation for 3 consecutive days to avoid non-specific activation of the LC neurons (Figure 2A). During this habituation session, room air was released into the chamber at a flow rate of 1.8L/min, in order to familiarize the animals with the gas flow and the sound of the air jet, reducing neophobic reactions. On the fourth day, the animals were then exposed to CO2-enriched air (20% CO2, 21% O2, N2 balance for 15 minutes). After exposure to the gaseous mixture, each animal remained in the experimental chamber for 60 minutes in ambient air in order to express c-Fos protein. Each animal was then deeply anesthetized with isoflurane, perfused, and its brain removed and prepared for immunohistochemistry (Figure 2A).
Following immunohistochemical staining for the c-Fos protein, photomicrographs were captured at a magnification of 20x to count the number of cells exhibiting double staining for TH and c-Fos, as well as those solely stained for TH, bilaterally. The analysis of the sections containing the LC consisted of subdividing the region into 4 main portions as previously defined by Ginovart et al. (1996): Caudal (bregma: −5.78 mm), Medial – Caudal (bregma: −5.65 mm), Medial – Rostral (bregma: −5.55 mm) and Rostral (bregma: −5.38 mm) (Figure 1). In addition to these, in the medial regions, there was also a subdivision into dorsal, medial and ventral (Dorso–ventral division). The analysis was performed based on the percentage of double-labeled cells in each subregion for each group for both sexes. The double-labeled (TH^+^ and c-Fos^+^) number of cells sampled in the LC was the sum of cells counted on the left and right sides. For the analysis, we used the mean of three different sections of each subdivision, along the rostrocaudal axis. Given the close proximity of the LC to the peri-LC region, the analysis extended to include c-fos^+^ neurons within this vicinity. The definition of peri-LC was established in accordance with the parameters outlined in the study by Luskin et al., (2022).
Animals were placed in the experimental chamber (a cylindrical chamber, 18 cm in diameter x 18 cm in height) with normocapnic air (flow of 1.8L/min) for 45 min for habituation. After the end of the habituation period, VE control measurements were performed at 15 minutes (Figure 3A). The animals were then subjected to 20% CO2 for 15 min, and at the end of the stimulus, a new VE measurement was performed. After the challenge, the animals were kept in room air for another 15 min when the last VE measurement was performed. For the VO2 measurements, the air in the chamber was sampled throughout the experiment, except during the VE recordings, when the chamber was sealed. During this period, the air passed directly from the aquarium pump (when in normocapnia) or the gas mixer (when in hypercapnia) to the analyzer, making it possible to compare the oxygen concentration of the chamber with that of the incoming air in the experimental apparatus.
To analyze panic attack behavioral responses, the experiments were recorded by two cameras (Logitech, USA) strategically positioned beside and above the experimental chamber, enabling the recording of the animal and its movements. The escape response analyzed included the number of jumps, the number of runs (running) and freezing (total lack of movement except breathing) (Figure 3A). All data were analyzed by two trained observers who were blinded to the treatment group assignments.
After the CO2 challenge, forebrain and brainstem monoamine concentrations were evaluated to assess sex- and genotype-specific alterations in catecholaminergic due to CO2 exposure. To this end, male and female animals of each genotype were deeply anesthetized with isoflurane and euthanized for brain extraction. The brains were frozen for the HPLC-ED analysis (Figure 5A).
All the results are reported as the mean ± SD. The data were tested for normality of deviation (Cramer Von-Mises criterion) and homoscedasticity (Levene test). To measure c-Fos in LC-NA neurons, bright-field immunohistochemistry photomicrographs of DAB-stained LC sections were acquired at 40× using a microscope (Zeiss, Axio Image Z2, Baden-Württemberg, Germany) using the LAS image acquisition software. The analysis was based on the quantification of immunoreactive c-Fos^+^ TH^+^ cells in the LC subdivisions using a computerized image analysis system (ImageJ). For the assessment of c-Fos expression in LC neurons, quantification was conducted on every fourth 40-μm coronal section containing the LC and included at least three sections from 5.38 to 5.78 mm caudal to bregma. An experimenter blind to the treatment group performed the quantification. For the peri-LC region the same coordinates were considered.
All respiratory parameter analyses were conducted using approximately 2-minute recording intervals, while the behavioral parameters were based on the 7 minutes of 20% CO2 exposure.
The number of c-Fos/TH-ir neurons was compared using a two-way ANOVA by Tukey’s post hoc test. The respiratory, metabolic and behavior variables, as well as the neurotransmitter concentrations, were compared between the three groups and two sexes by two-way ANOVA, followed by Tukey’s post hoc test. P<0.05 was considered statistically significant.
All statistical analyses performed are shown in Supplementary File (Tables S1–S9; S11).
Considering the CO2 sensitivity of the LC and its association with anxiety-related behaviors, our objective was to investigate whether severe hypercapnia elicits distinct patterns of activation of NA neurons in the rostral region of the LC in wild-type male and female mice. We found that males exposed to 20% CO2 exhibited a significantly higher number of activated TH^+^ neurons in comparison to males in the control group (0% CO2, F(1,12)= 22.081; p= 0.001, Figure 2B) and females exposed to a 20% CO2 F(1,12)= 13.799; p= 0.003). Conversely, this differential activation did not occur in females.
Regarding the mid-rostral LC region, exposure to a 20% CO2 environment induces a heightened activation of noradrenergic cells only in males (Figure 2C). This activation primarily manifests in the medial and ventral regions of the rostral LC division (medial: F(1,12)= 9.644; p= 0.009 compared to males exposed to 0% CO2, and F(1,12)= 12.503; p= 0.004 compared to females; F(1,12)= 7,981; p= 0.015 compared to males in 0% CO2 and F(1,12)= 8.152; p= 0.014 compared to females). The dorsal subregion of the medial-rostral LC division exhibit differences between male and female in the number of activated NA-neurons (F(1,12)= 6.163; p=0.029).
Regarding the mid-caudal LC subdivided into dorsal, medial and ventral, we observed that 20% CO2 promotes greater activation of noradrenergic cells only in males (Figure 2D). This activation occurred in all subdivisions (dorsal: F(1,12)= 8.035; p= 0.001, compared to males in 0% CO2 and F(1,12)= 7.175; p= 0.020 compared to females; medial: F(1,12)= 9.611; p= 0.009 compared to males in 0 % CO2 and F(1,12)= 5.037; p= 0.044 compared to females; ventral: F(1,12)= 4.187; p= 0.053 compared to males in 0% CO2 and F(1,12)= 6.823; p= 0.023 compared to females).
Finally, both males and females exposed to an environment with 20% CO2 show no difference in the caudal LC region (Figure 2E), when compared to normocapnic values, despite a trend towards an increase in neurons positive for the c-Fos protein presented by males. Our findings demonstrate that severe hypercapnia activates LC-NA neurons in males, while no activation was observed in females.
Regarding the peri-LC region, no difference in c-fos expression was observed in this region (Table S10, S11).
Under room air conditions, Dbh^LC-null^ and Dbh^hypo^ male mice had a lower VE than the wild-type control group (Dbh^wt/wt^) (F(2,30)= 4.151; p= 0.026 - Figure 4A). Dbh^LC-null^ females exhibited a reduced VT compared to wild-type controls (F(2,30)= 5.574; p= 0.009) with no significant difference from Dbh^hypo^. Furthermore, male wild-type controls showed a higher fR compared to female controls (F(2,30)= 4.648; p= 0.017). However, for the remaining variables, there were no significant differences observed based on genotype or sex (Figure 3B). Dbh^LC-null^ male mice showed a higher VO2 compared to control animals (F(2,30)= 3.794; p= 0.034) (Figure 3B). Both Dbh^LC-null^ and Dbh^hypo^ males present a lower VE/VO2 compared to wild type (F(2,30)= 6.177; p= 0.006). No significant differences in the remaining variables were observed with respect to genotype or sex (Figure 3B). Collectively, our results demonstrate that Dbh mutation in males causes hypoventilation.
20% CO2 caused an increase in VE all male and female mice examined (F(1,60)= 307.566; p< 0.0001) because of increases in VT (F(1,60)= 218.746; p< 0.0001) and fR (F(1,60)= 64.105; p< 0.0001) (Figure 3C). Dbh^LC-null^ male mice presented a lower fR compared to wild-type males (F(2,30) = 3,918; p= 0.031) and Dbh^LC-null^ females (F(1,30)= 5.157; p= 0.030), while Dbh^LC-null^ females showed a higher VT compared to Dbh^hypo^ and wild-type females (F(2,30)= 3.876; p= 0.032) and Dbh^LC-null^ males (F(1,30)= 6.188; p= 0.019). Further, Dbh^hypo^ females showed a higher CO2-induced hyperpnea compared to wild-type females.
Severe hypercapnia induced a reduction in both VO2 (F1,60)= 61.341; p< 0.001) and Tb (F(1,60)= 75.853; p<0.001) across male and female groups (Figure 3C) and an increase in VE/VO2 (F(1,60)= 75.688; p<0.001). The genotype did not interfere with metabolic and thermal response to severe CO2 exposure. Together, our findings demonstrate that LC-NA neurons impact breathing control differentially in males and females under severe CO2 challenge, promoting an excitatory modulation in fR in males and an inhibitory effect on VT in females.
Our next goal was to ascertain whether the LC-NA neurons are involved in panicogenic responses induced by CO2, given the established role of the LC-NA neurons in anxiety (Morris et al., 2020). Under normocapnia, no panicogenic response was observed in any group (Figure 4A). Exposure to 20% CO2 promoted an increase in jumping in wild-type mice of both sexes (F(1,78)= 31.283; p<0.001). However, this increase was smaller in females (Figure 4B). Dbh^LC-null^ animals had a blunted jumping response compared to Dbh^wt/wt^ (F(2.18)= 11.497; p= 0.001 for males and F(2.21)= 5.328; p= 0.0013 for females). Dbh^LC-null^ males and females showed a blunted running response compared to wild-type animals (F(2.18= 5.539; p= 0.013 for males and F(2.21)= 3.995; p= 0.034 for females) (Video S1). Dbh^LC-null^ males also differed from Dbh^hypo^. Females of all groups presented a higher running response compared to males (F(1,39)= 23.770; p< 0.001). In addition, Dbh^LC-null^ males exhibited a longer duration of freezing behavior than wild type, Dbh^hypo^ and Dbh^LC-null^ females (F(2.78)= 8.387; p< 0.001 - Figure 4B).
To assess the concentration of monoamines in the brain of Dbh mutant and control mice after 20% CO2 exposure, we measured noradrenaline (NA), dopamine (DA), and 3,4-dihydroxyphenylacetic acid (DOPAC) levels in dissected tissue from the forebrain and brainstem using HPLC.
Regarding forebrain, Dbh^LC-null^ and Dbh^hypo^ mice of both sexes had a lower NA concentration compared to wild type (males: F(2,19)= 10.257; p = 0.001; F(2,18)= 20.530; p<0.0001), confirming that the mutation was effective. No differences were observed in the other variables in terms of sex or treatment (Figure 5B).
As to brainstem, Dbh^LC-null^ and Dbh^hypo^ males presented higher DA concentration compared to Dbh^wt/wt^ (F(2,19)= 6.057; p = 0.009). Dbh^LC-null^ and Dbh^hypo^ females showed a higher DOPAC concentration, the main DA metabolite, when compared to wild-type controls (F (2,18) = 18.536; p< 0.001). No differences were observed in terms of genotype or sex for the remaining variables. (Figure 5C).
It is well established that LC-NA neurons are CO2/pH chemosensors and are involved in hypercanic hyperventilatory response in rats (Gargaglioni et al., 2010), but their involvement in breathing and behavioral regulation under severe CO2 exposure in mice remains less understood. In the current study, we found that severe hypercapnia activates LC-NA neurons and that males presented a higher activation compared to females. Further, we investigated the ventilatory, metabolic and behavioral responses exhibited by mice with mutations in the dopamine β-hydroxylase gene, encoding the enzyme responsible for converting dopamine into noradrenaline. We used Dbh^hypo^ mice, which have reduced Dbh expression in noradrenergic neurons, and an LC-specific knockout, Dbh^LC-null^, which has no Dbh expression in LC-NA neurons and reduced expression in other noradrenergic neurons (Wilson et al., 2023). Importantly, our study reveals that Dbh^LC-null^ mice of both sexes display a blunted escape response when exposed to elevated levels of CO2, indicating a role of the locus coeruleus noradrenergic system in the panicogenic response triggered by severe hypercapnia.
In our study, we observed significant sex-related differences in c-Fos activation specifically within the LC-NA neuron population. CO2 inhalation activates chemosensors located both in the periphery and within the brain (Feldman et al., 2013). This activation subsequently leads to a vigorous increase in ventilation (hyperventilation). The CO2 chemosensory circuit also plays a significant role in the pathophysiology of panic disorder (Nardi et al., 2008). When chemosensitivity to CO2 and/or its responsiveness becomes overly heightened, it gives rise to physiological and emotional responses commonly associated with fear and panic. In our observations, we noted that severe hypercapnia activated NA neurons in all subregions of the LC in males, mirroring the findings of a prior study involving rats (Johnson et al., 2005). The observed activation was specifically attributed to LC neurons, as no discernible differences were noted within the peri-LC region. In contrast, females did not exhibit LC activation in any of the analyzed subregions. These findings are consistent with the fact that females respond less to CO2 compared to males as can be observed in CO2-induced panicogenic behavior (Figure 4). Thus, the sex-specific CO2 stimulation pattern in the LC might contribute to the differences in behavior observed in males and females. This sex-based difference in LC neuronal activation may stem from variations in the distribution of sex steroid hormone receptors in this region and circulating levels of sex hormones. Previous study indicates that male and ovarectomized (OVX) rats exhibit similar increases in cerebral blood flow (CBF) following CO2 exposure (Ances et al., 2001). However, non-OVX female rats displayed CBF similar responses to pre-hypercapnia levels, significantly lower compared to males and OVX females. Consequently, the authors propose that estrogen contributes to reduced responsiveness in heightened CBF induced by hypercapnia. Hence, there is a possibility that estrogen exerts its influence on the LC region, inhibiting the activation of NA neurons. Further investigations are warranted to explore into this matter.
We observed that Dbh^LC-null^ and Dbh^hypo^ males hypoventilated during normocapnia promoted by a decrease in VE in both groups and an increase in VO2 in Dbh^LC-null^. Females showed a decreased VT in room air conditions, but, unlike males, no change in VE was observed. Previous studies have demonstrated that LC neurons exhibit a respiratory-related activity, i.e., they have direct access to information about the timing of the respiratory output from the medullary respiratory centers (Oyamada et al., 1998, 1999; Andrzejewski et al., 2001). In this regard, selective lesion of the LC in adult rats using 6-OHDA (a toxin that selectively eliminates catecholaminergic neurons) did not change basal ventilation in male rats, suggesting that noradrenergic neurons located in the LC play no role in respiratory control under resting conditions in adults (Biancardi et al., 2008). Further, the injection of a potent toxin conjugate, SP-SAP, in the LC of adult rats for killing neurons expressing the neurokinin-1 (NK-1) receptor did not alter adult breathing under basal conditions (De Carvalho et al., 2010). Different from these data, we observed that Dbh^hypo^ genotype and Dbh^LC-null^ male mice presented hypoventilation, due to a lower VE and a higher VO2. Therefore, it seems that NA from LC and the brainstem, in general, is important to maintain the respiratory drive under room air conditions. Since Dbh^hypo^ and Dbh^LC-null^ mice have statistically significant differences, then the LC is implicated, but the other noradrenergic centers may still explain the difference from chemical lesioning. The differences observed among the current data and previous studies might be related to different techniques (knockout animals x lesion techniques) or different species (mice x rats). Lesion approaches may promote neuroinflammatory responses that might interfere with the physiological responses. In addition, lesions can also indirectly affect neighboring cells and disturb brain homeostasis (Barut et al., 2022).
Regarding the ventilatory response to 20% CO2, Dbh^LC-null^ males showed a lower fR, while Dbh^LC-null^ females presented an increase in VT when compared to wild-type and Dbh^hypo^ females and Dbh^LC-null^ males. There is substantial evidence demonstrating that the LC acts as a central chemoreceptor in rats, being intrinsically sensitive to CO2/pH (Elam et al., 1981; Filosa et al., 2002; Gargaglioni et al., 2010; Imber et al., 2014; Pineda and Aghajanian, 1997). In fact, this region plays an important role in respiratory control, since >85% of LC neurons are sensitive to CO2/pH (Johnson et al., 2008). Studies using 7% CO2 combined with LC lesions demonstrated a decrease in ventilatory response compared to controls. Biancardi (2008) demonstrated a 64% drop in response after injury to approximately 80% of LC neurons. These data demonstrate that at a moderate concentration of CO2, the LC plays a key role in the chemosensitive response, at least in rats, but at high concentrations, as performed in the present study (20%), other chemosensitive areas seem to be acting on the ventilatory response, since no change in respiratory equivalent (VE/VO2) was observed. As highlighted in the preceding paragraph, in this study, aside from the elevated CO2 concentration, we employed mice as subjects, a factor that may have influenced the observed variations. Even though Dbh^LC-null^ males presented a lower fR, no difference in VE was observed, possibly due to a compensation of VT, which presented a trend to increase. Moreover, our observations reveal that Dbh^LC-null^ females employ a similar ventilatory response mechanism to high CO2 concentrations as males, characterized by an elevation in VT while exhibiting no discernible alterations in the respiratory equivalent. As demonstrated in Figure 6, wild-type females had a lower concentration of DOPAC in the brainstem than Dbh^LC-null^ and Dbh^LC-hypo^ animals, which indicates an increase of DA being metabolized by mitochondrial monoamine oxidase (MAO) to DOPAC under CO2 exposure. On the other hand, mutant males do not present such a difference. It is reasonable to affirm that females may compensate the lack of NA and it is presenting a compensatory increase in the metabolization of DA. Hence, the differences noted between males and females may be attributed to the levels of DOPAC in the brainstem.
Concerning the behavioral data, it is possible to observe a divergence in behavioral patterns between wild-type males and females. Specifically, males exhibit a higher frequency of jumps compared to females, whereas females engage in more frequent running behaviors than males. Moreover, our results reveal that Dbh^LC-null^ males exhibit extended periods of freezing behavior during CO2 exposure and display reduced levels of jumping and running in comparison to wild type. In mice, inhalation of 20% CO2 causes avoidance and freezing (D’amato et al., 2011; Leibold et al., 2016; Taugher et al., 2014; Vollmer et al., 2016; Ziemann et al., 2009) and endocrine changes, which suggest that the animals are experiencing a highly aversive situation, which resembles the triggering of a panic attack (Spiacci et al., 2018). Klein (1993) proposed that every individual has an alarm system that detects metabolic threats such as increased CO2 concentrations and reduced blood pH. These changes alert individuals of impending asphyxia by activating their suffocation detection system, causing an emotional and behavioral response such as acute panic. A dysfunctional system, however, can be mistakenly activated without the presence of a real threat, eventually triggering a panic attack.
In this context, the LC stands out for being a known chemosensitive site (Elam et al., 1981; Filosa et al., 2002; Gargaglioni et al., 2010; Imber et al., 2014; Pineda and Aghajanian, 1997) and for sending projections to several areas responsible for respiratory control and places responsible for triggering panic-related behaviors, fear and anxiety behaviors such as the amygdala (Mccall, et.al., 2017; Giustino et.al., 2020). It is known that the amygdala is associated with conditioned aversive stimuli, and is responsible for generating fight, flight, and fear behaviors, sending several projections to cortical areas such as the paraqueductal gray matter, dorsal hippocampus, among others. (Fanselow, 1994; Ledoux, 1996; Mcnaughton and Corr, 2004; Mobbs et al., 2009; Raybuck and Lattal, 2011). Thus, inhibiting the production of the main neurotransmitter in the LC can cause changes in the behavioral and respiratory responses generated in the face of a panicogenic stimulus.
Finally, the behavioral profile presented by Dbh^LC-null^ animals is characterized by extended freezing periods, little locomotion and a notably small number of jumps. In contrast, wild-type animals show an opposite pattern marked by extensive locomotor activity, a high frequency of jumps and minimal freezing time. These findings are in line with other studies that demonstrate a high occurrence of jumps in response to 20% CO2 (Spiaci, 2018, Taugher, 2019). These observations suggest that noradrenaline originating from the LC has an important role in the characterization of jumping and freezing behaviors. In their study, McCall et al. (2015) observed that the increased tonic activity of LC noradrenergic neurons generated aversive behavioral responses and typical panic behaviors. It was also observed that the LC-NA innervates the basolateral amygdala (BLA) and that the activation of this circuit generates the same responses observed previously (McCall et al., 2017). Since BLA receives NA released from LC axons during episodes of acute stress (Quirarte et al., 1998, Valentino and Van Bockstaele, 2008), it is plausible that the LC exerts a fine but indirect control of the system’s fight and flight responses, since known anxiogenic and panicogenic sites such as the ventral hippocampus and the central portion of the amygdala are activated by the BLA.
In summary, our study reveals that when exposed to severe CO2 levels, male LC-NA neurons exhibit activation, while female counterparts do not, highlighting a notable disparity in CO2 response between the sexes. Additionally, the release of noradrenaline from the LC is important for the tonic excitatory control of ventilation in male mice, but not in females. Under CO2 exposure, LC-NA neurons promote an excitatory modulation in fR in males and an inhibitory effect on VT in females. Further, our data provide emerging insight into the role of LC-NA neurons in panicogenic CO2 responses demonstrating the important role of this region in the active escape responses under severe hypercapnia. Together, our findings suggest that embryonic disruption of LC-NA has selective, sex-specific effects on ventilatory and behavioral responses to CO2.