Why do opiates cause respiratory depression

For a mouse that does not recover, the red vertical arrow indicates survival time when breathing stops i. A bracketing approach was used to construct a dose—response curve with a minimal number of mice. These traits are largely uncorrelated. Recovery time and survival time are mutually exclusive traits and each mouse can either recover or fail to recover, not both.

These values indicate that the traits are heritable and also amenable to genetic mapping. Strain- and sex-specific effect of morphine on respiratory sensitivity. The traits of recovery time or survival time are censored such that a mouse does not appear in both graphs as each mouse displays only one of these two phenotypes.

Empty bars indicate that no mice fell into this category i. Scatterplot of percent respiratory depression vs recovery time top right panel showing a weak correlation of R 2 of 0. Published data reveal a variety of morphine LD 50 values for mice. Yoburn et al. Using this approach, survival curves by strain and sex were established Fig.

The morphine LD 50 ranged from There was not a consistent sex bias up or down across all strains; instead, some LD 50 values were higher in females than males e.

The morphine LD 50 by strain and sex. The morphine LD 50 was determined for each of the eight founder strains and sex using at least six mice in each group and at least three doses of morphine, with at least two doses flanking the LD A Logistic 2-paramater survival curves separated by strain and sex are shown at the top and composites of all strains separated by sex are shown at the bottom.

JMP To find genetic loci that influence OIRD, quantitative trait loci QTL mapping was performed on the respiratory phenotypes using the high diversity, high precision, DO mouse population. Of the DO mice entered into the study, 83 females, males recovered and 67 females, 40 males did not recover. The quantitative metrics for respiratory response to morphine, including respiratory depression, recovery time and survival time, all show a continuum of phenotypic diversity Fig.

Respiratory response to morphine in Diversity Outbred mice. The distribution of respiratory responses is shown for respiratory depression top panel , recovery time middle panel and survival time bottom panel. A significant QTL was identified for respiratory depression, but no genome-wide significant QTLs were detected for recovery time or survival time using these sample sizes. For the respiratory depression trait, we identified a LOD 9. The QTL is called Rdro1 respiratory depression, response to opioids 1.

QTL mapping of respiratory depression in DO mice. B Allele effect plot of the LOD 9. Genetic mapping studies are used to identify regions of interest containing variants that influence complex traits. To identify the relevant genes involved in complex trait regulatory mechanisms, there must be evidence of genetic polymorphisms segregating in the population that either influence protein structure or gene expression and evidence of a biological mechanism of action connecting them to the trait, such as expression in a trait-relevant tissue.

We identified a While these SNPs remain candidates for regulation of the respiratory depression phenotype, we focused on coding SNPs because their impact is more readily predictable. None of the coding SNPs are the type with the most deleterious effects, such as a stop loss, stop gain or coding region insertion frameshift. Two of the three changes serine to threonine at amino acid in Kmt2c and asparagine to aspartic acid at amino acid 29 in Speer4a occurred in residues that are not conserved across species.

This amino acid is located within a functional domain that is conserved in vertebrates Fig. Indeed, this change places the hydrophobic residue, which are generally buried internally, onto the surface of the protein. The 3D protein structure analysis Fig. A Multi-species alignment of the Ricin B lectin domain of GALNT11 showing that the SL mutation occurs in a domain conserved across multiple vertebrates, including humans, mice, rats, Xenopus , and zebrafish.

The 3D structure rendered showing secondary structure as a cartoon type with coloring as a rainbow from N- to C-terminus. In this study, we found heritable strain differences in the quantitative metrics of respiratory response to morphine, including respiratory depression, recovery time and survival time, using an advanced, high-throughput, behavioral phenotyping protocol.

We further identified genomic loci involved in morphine-induced respiratory depression using an unbiased genetic approach. Mapping these traits in the DO mice and evaluation of sequence variants and protein structure, followed by integrative functional genomic analysis in GeneWeaver, has allowed us to implicate Galnt11 as a candidate gene for respiratory depression in response to morphine. We identified specific inbred strains of mice that were more sensitive to morphine than other inbred strains of mice.

The traits of respiratory depression, recovery time and survival time were all shown to have a high degree of heritability. In determining our probe dose for the outbred population, we observed that the LD 50 for morphine differed by four-fold between these eight parental strains harboring 45 million SNPs, or an equivalent genetic variation as found in the human population.

The traits of respiratory depression, recovery time and survival time were largely independent traits, as seen by their lack of correlation. Strains such as AJ, which had the lowest LD 50 for both males and females, did not demonstrate the highest degree of respiratory depression, suggesting that factors other than respiratory depression may play a role in opioid overdose.

The lack of correlation between percent respiratory depression and survival time suggest other mechanisms of death in conjunction with OIRD. Dolinak suggests that other baseline characteristics, such as obstructive sleep apnea, obesity, heart disease and lung disease, make an individual more susceptible to opioid toxicity The presence of the alleles segregating in the DO population are encouraging for finding additional QTLs related to recovery time and survival time in larger cohorts and possibly using alternatative opioids.

The mechanisms underlying sensitivity to morphine and fentanyl are known to differ in many respects and this same work should be performed for the more potent fentanyl.

Fentanyl has similarities to morphine with respect to the recruitment of intracellular signaling mechanisms but there are also key differences.

Only one study has looked at human opioid overdose risk, specifically by scoring overdose status and determining the number of times that medical treatment was needed in European American populations Human genes have thus been mapped to opioid use, opioid dependence and opioid overdose susceptibility but human studies are not able to assess opioid-induced respiratory depression, specifically the LD 50 of an opioid.

Animal studies have allowed us the opportunity to assess the LD 50 of a drug in a variety of genetic backgrounds and then map those sources of variation. These types of controlled exposure experiments cannot be conducted in humans for which exquisite control of environment is not feasible and prior exposure history is unknown.

Our genetic approach of QTL mapping in the DO mouse population has allowed us to identify a genomic region containing no genes previously known to function in opioid pharmacodynamics or pharmacokinetic processes, or implicated in OUD. The genetically diverse structure of this population allows for the identification of narrow genomic intervals often with very few candidate genes. This approach of using advanced mouse populations together with integrative functional genomics has been useful for the prioritization of candidate genes in a variety of different disciplines 62 , 63 , The identification of Galnt11 as functioning within the morphine respiratory response reveals a potential new target for therapeutic development.

GALNT11 is an N -acetylgalactosaminyltransferase that initiates O-linked glycosylation whereby an N -acetyl- d -galactosamine residue is transferred to a serine or threonine residue on the target protein. The lectin domain of GALNT11 is the portion that functions to recognize partially glycosylated substrates and direct the glycosylation at nearby sites.

This type of post-translational modification controls many phamacokinetic and pharmacodynamic processes as well as the regulation of delta opioid receptor OPRD1 membrane insertions as O-linked glycosylation is required for proper export of OPRD1 from the ER O-linked glycosylation is also required for opioid binding peptides, increasing their ability to cross the blood brain barrier The integrative functional analysis in GeneWeaver identified Hs6st2 36 , Fn1 37 , Lrp1 36 , and Sdc4 38 as glycosylation targets of Galnt Our findings demonstrate the initial mapping of a locus involved in OIRD in mice, for which the likely candidates do not act via the opioid receptor, thereby providing a potential new target for remedial measures.

Although it is through mouse genetic variation that we identified this gene, it should be noted that this gene or its glycosylation targets need not vary in humans to be a viable target mechanism for therapeutic discovery and development.

Characterization of the role of Galnt11 and its variants along with other viable candidates will resolve the mechanism further, and continued mapping studies in larger populations will enable detection of additional loci for various aspects of the opioid-induced respiratory response.

These findings suggest that phenotypic and genetic variation in the laboratory mouse provides a useful discovery tool for identification of previously unknown biological mechanisms of OIRD. All mice were acquired from The Jackson Laboratory JAX and were housed in duplex polycarbonate cages and maintained in a climate-controlled room under a standard light—dark cycle lights on at 0, h.

Bedding was changed weekly and mice had free access to acidified water throughout the study. Louis, MO. A Nestlet and Shepherd Shack were provided in each cage for enrichment. Mice were housed in same sex groups of three to five mice per cage. A mouse plethysmography chamber was built consisting of a ml plexiglass chamber with ports for air supply and pressure measurement, and end openings outfitted with rubber gaskets to create an airtight seal after closing.

A piezo film sensor was sealed into the bottom of the chamber for data collection with PiezoSleep software, with piezo signal sampling set at Hz.

A differential pressure transducer Biopac measures the pressure difference between the plethysmography chamber and reference chamber sampled at Hz. Signal alignment was done through a simple correlation, which typically indicates time differences of several seconds. The program graphically displays an overlay of the piezo and plethysmography signals for easy visual confirmation. A graphical breath rate overlay allows navigation to intervals of signal disagreement to inspect the signals at these areas.

Not all strains received all doses but each strain received at least three doses such that two flanked one above, one below the LD Individual testing is necessary due to the known enhanced lethality of cage mates during morphine exposure, which has been shown to affect survival The mice had access to food and water ad libitum while in the chamber. The room was maintained on a h light:dark cycle. They remained in their chambers undisturbed until 24 h after injection.

Whenever possible, complete balanced cohorts of the eight strains and both sexes were run during each of nine replicates of the experiment. The data acquisition computer, food and water were checked daily; otherwise, the mice remained undisturbed. Breath rates were estimated from 4-s intervals in which animal activity dropped low i.

The respiratory rate baseline consisted of the average respiratory rate over the first 24 h, which included both sleep and wake periods. Respiratory rate was then measured in the same way after injection of opioid. These measures were then used to determine thresholds for obtaining the recovery time respiratory rate returns to baseline, see Fig. This dose was determined as the average LD 50 dose across the eight strains and two sexes, 16 samples.

To test for difference in the respiratory depression, recovery time and survival time across the strains and sexes a linear model was fit, the full model was:. In all cases, the full model was fit and reduced by dropping non-significant interactions followed by main effects. The LD 50 was calculated using the drc 3. A logistic regression model was fit, and a goodness of fit test based upon Bates and Watts 72 performed.

In addition, a regression model assuming equal LD 50 across strains was compared by chi square to an LD 50 assumed different across strains. To graph the data a non-linear 2-parameter model was fit in JMP Tail samples were collected at the conclusion of the experiment and all mice were genotyped using the Giga-MUGA genotyping array NeoGene. Genotypes were imputed to a 69 K grid to allow for equal representation across the genome. Genotype probabilities were calculated according to the founder genotypes and then converted to allele probabilities.

We then interpolated allele probabilities into a grid of 69, evenly-spaced genetic intervals Sex and date of test were included as additive covariates. The significance thresholds were determined for each trait by permutation mapping The confidence interval around the peak makers was determined using Bayesian support intervals.

To determine the percent of variation accounted for by the QTL the mice were classified at the variant into one of eight states based on genotype probabilities of mice at that locus. Following this, a one-way ANOVA was fit to ascertain the strain variation relative to total variation towards estimating heritable variation at that locus. At each SNP location the eight allele state probabilities are collapsed to two state SNP probabilities and the Cox proportional hazards regression was performed by coxph function in the survival 3.

Based on the output of the log base e likelihood for the null model and for the alternative model with covariates and genotype probabilities , we took the difference of both log likelihoods and then divided by ln 10 to convert the results into the LOD scale.

Next we identified those that were within protein coding region that were most deleterious. Differential coding sequence non-synonymous amino acid substitution SNPs Cn that differed between the high and low allele groups were identified. The gene sets were overlapped using the Jaccard similarity and GeneSet graph tools In order to determine if the Cn SNPS were in areas of evolutionary conservation we aligned the sequence of several species.

The Clustal Omega program was used with default parameters The transition matrix is Gonnet, gap opening penalty of six bits, gap extension of one bit. Clustal-Omega uses the HHalign algorithm and its default settings as its core alignment engine In order to assess the functional sufficiency of Galnt11 as a candidate gene the literature was searched to identify genome-wide studies characterizing glycosylation targets of GALTN11, one study was identified 33 and these genes were added to the GeneWeaver Database GS Boscarino, J.

Risk factors for drug dependence among out-patients on opioid therapy in a large US health-care system. Addiction , — PubMed Google Scholar. Cicero, T. Increased use of heroin as an initiating opioid of abuse: Further considerations and policy implications. Schmid, C. Bias factor and therapeutic window correlate to predict safer opioid analgesics. Cell , Tomassoni, A. Multiple fentanyl overdoses—New Haven, Connecticut, June 23, Google Scholar. Comer, S. Fentanyl: Receptor pharmacology, abuse potential, and implications for treatment.

May, W. Morphine has latent deleterious effects on the ventilatory responses to a hypoxic—hypercapnic challenge. Open J. Pattinson, K. Opioids and the control of respiration. Frischknecht, H. Opioids and behavior: Genetic aspects. Experientia 44 , — Baran, A. Opiate receptors in mice: Genetic differences.

Life Sci. Shigeta, Y. Pharmacogenetics Genomics 18 , — Juni, A. Nociception increases during opioid infusion in opioid receptor triple knock-out mice. Rhythmic activity was normalized to the first control recording for dose response curves.

The final libraries were sequenced on HiSeq For analysis, sequencing reads were processed using the 10x Genomics Cell Ranger v. A total of cells were sequenced. Further analysis was performed using Seurat v2. Cells with less than genes were removed from the dataset. Data was LogNormalized and scaled at 1e4. Highly variable genes were identified and used for principal component analysis. FindAllMarkers and violin plots of known cell type markers were used to identify each cluster.

Summary data generated in this study are included as a supplemental supporting file. In the interests of transparency, eLife publishes the most substantive revision requests and the accompanying author responses. This is a technically sophisticated study employing mouse transgenic approaches, viral vector-based neuronal targeting, in vivo behavioral analyses, single cell transcriptomics, and in vitro electrophysiological measurements to explore the mechanisms underlying opioid respiratory depression.

The tour-de-force of impressive tools and the clear presentation make this paper especially appealing to a wide audience beyond those who study respiration. Thank you for submitting your article "Opioids depress breathing through two small brainstem sites" 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 Eve Marder as the Senior Editor.

The following individuals involved in review of your submission have agreed to reveal their identity: Gaspard Montandon Reviewer 2 ; Jeffrey C. Smith Reviewer 3. The reviewers have discussed the reviews with one another and the Reviewing Editor has drafted this decision to help you prepare a revised submission.

The experimental approaches employed are novel, providing a number of important results that more clearly define the critical brainstem sites and molecular phenotypes of neurons involved in OIRD. The paper is well written and the figures and supplementary material very nicely illustrate the essential experimental results.

However, there are several concerns that need to be addressed. Yes, it would be nice if there were a simple gold standard. However, this is not the reality. Gold standard for OIRD should come from a consensus of studies showing similar results. Also the "gold standard" proposed here is based upon flow which cannot be reliably measured with whole-body plethysmography.

So don't state that you are introducing a gold standard. OIRD is very state dependent and shows huge interindividual variability. Trying to imply that the respiratory rhythm is simply depressed is an oversimplification. Just look at the huge variability in Te Figure 1H. Plethysmograph recordings are extremely difficult to evaluate, because animals behave, morphine will be sedating before measuring the OIRD, this will not be a static response. We don't understand the definition of inspiratory time, expiratory time, and pause.

It seems in Figure 1G that flow doesn't cross the zero at the end of expiration as shown in the figure, but rather at the end of the pause.

Is the "pause" only a longer expiration? How was this pause detected or defined? If those pauses are part of expiration, then expiratory time is also increase by morphine. How was the criterion for pause of 0. Knowing that flow is highly variable from one animal to the other it is very likely an unreliable threshold.

Along the same lines: The authors respiratory assessments in freely-behaving rodents limited for a few reasons: a using an arbitrary threshold based on flow to define the end of expiration is not reliable. Flow can change from one animal to another, so the duration of their "pause" may differ depending upon the initial baseline of the animal. Unless the right formula is used, it is very difficult to evaluate tidal volume in rodents. In addition, mL cannot be a correct measurement in rodents without normalizing the value according to body weight.

You need to carefully consider these caveats. It is well know that sedation by opioids may affect the severity of respiratory depression so arousal levels before morphine may be an issue. Please discuss this important caveat. Under hypercapnia, arousal levels may be severely disrupted and it has been shown that arousal levels are tightly linked to respiratory depression.

Hypercapnia will also affect the opiate sensitivity, and this effect will not be uniform in different areas. Carefully discuss this important caveat, and don't imply that this is an advantage, other than the fact that it regularizes the rhythm.

Figure 3—figure supplement 1. It is possible that the inflammation induced by AAV, and the damage done by the microinjection may change the response to opioids. We recommend that this should be verified with proper controls. It would be important to emphasize in the Discussion that tests need to be performed in the awake behaving mice in vivo with the transgenic lines utilized for the in vitro studies to determine if so few neurons are also responsible for OIRD in the adult system in vivo.

Thank you for resubmitting your article "Opioids depress breathing through two small brainstem sites" for consideration by eLife. Your article has been reviewed by two peer reviewers, and the evaluation has been overseen by Ronald Calabrese as the Senior Editor and Reviewing Editor. The following individuals involved in review of your submission have agreed to reveal their identity: Gaspard Montandon Reviewer 2 ; Jeffrey C Smith Reviewer 3.

This paper has been previously reviewed and requires some further revisions. All these revisions can be accomplished without any further experiments. Some new analysis in needed but the vast majority of the changes involve rewriting.

So it is clear that the authors have included the targeted areas with their stereotaxic injection approach. The authors could confirm the absence of MOR in knockout animals with Immunohistochemistry. They need to mention that these data are valid in animals that are exposed to CO2, and discuss the issues related to high CO2, low pH, hyperventilation induced by CO2, and reduced arousal, which may limit the role of the preBotC in opioid-induced respiratory depression.

The authors don't mention any papers in the discussion related to hypercapnia, medullary sites sensitive to CO2, arousal states, opioid receptor knockout animals etc. Definition of pause versus expiration. The definition of a pause according to the authors is when volume goes below 0.

According to Figure 1G, the last part of the pause with morphine seems to be above this threshold. This arbitrary threshold is concerning. The threshold is likely detecting different pauses in saline versus morphine considering that volume is considerably reduced by morphine. The use of an absolute value 0. It would be easier to use the simple definitions of inspiratory and expiratory times, which would be consistent with previous reports showing that expiratory duration is affected by opioids.

Also, we are not sure what the physiological relevance of a pause is. Is it controlled by different brainstem circuits? The long expirations are likely due to a combination of hypercapnia and hypoxia with morphine. There are many studies investigating this topic that could be discussed in the Discussion. Dahan et al. We use this second threshold to distinguish Te from the pause. The specific details of the pause are discussed below.

The pause is an expiratory period that occurs after the initial large-amplitude expiratory pulse wherein the majority of tidal volume has been expelled.

The pause is characterized by very low airflow and ends with a final brief expulsion of air. Since this phenomenon is being described in hypercapnia, we interpret this terminating expiratory airflow to be "active expiration" Pisanski and Pagliardini, The low airflow pause period can vary in length from tens to hundreds of milliseconds.

Additionally, we have added a description in the Materials and methods and Figure 1—figure supplement 2 to clarify how the airflow threshold was chosen to define pause onset. The pause is an apneic period of low airflow which we believe represents a failure or delay to initiate a subsequent breath. This period is qualitatively defined as the point during expiration when expiratory airflow crosses below 0.

It is certainly true that the overall time between two inspirations does increases. However, the initial portion of expiration before our pause does not change in length. This is the data displayed in Figure 1H. Since the pause analysis was conducted on data from adult animals in hypercapnia, airflow recordings are very reproducible between animals and not "highly variable". Furthermore, in Figure 1—figure supplement 2, where the pause threshold is defined, the data in the scatter plot was generated from breaths from 15 different animals and all data points are intermixed and not clustering as individual animals.

Therefore, our pause threshold is chosen to be a reliable metric between animals. Along the same lines: The authors respiratory assessments in freely-behaving rodents limited for a few reasons: a Using an arbitrary threshold based on flow to define the end of expiration is not reliable. Please see explanations above and Figure 1—figure supplement 2. Briefly, data and pause threshold between animals is highly reproducible and data from 29 animals was used to define the pause threshold.

We appreciate the criticism that whole-body plethysmography inaccurately measures tidal volume and that a more correct estimate necessitates the incorporation of chamber humidity and temperature via the Drorbaugh and Fenn formula.

It is certainly important to normalize the airflow and tidal volume to body weight when animals of very different sizes are being compared. For example, this correction is particularly important when comparing developing rodents where lung size strongly correlates with size. However, we feel that this correction is not appropriate in our manuscript for four reasons:. These two recordings are performed within several days of each other i. All statistical tests are performed on normalized data.

However once mice reach their adult size, although their body weight can increase, their lung volume does not. This has been clearly demonstrated with serial computed tomography imaging of lung size Mitzner, Brown and Lee, See Author response image 1. This indicates that normalizing approximate airflow or tidal volume to weight, as a proxy for body volume, will negatively impact the accuracy of those estimates.

When airflows from the same animal, after injection of saline, are compared at the beginning and end of our studies, the airflows are indistinguishable despite some mice increasing in weight See Author response image 2 where specific breaths in hypercapnia are compared. In this case normalizing by weight would lead the experimenter to believe lung volume significantly decreases as animals increase in weight which would have no biological basis.

In summary, it is unlikely that either animal weight or size in adult mice is significantly biasing approximated respiratory airflow or tidal volume values, and therefore it would be inappropriate to normalize these values by weight. The plethysmograph was injected with 10mL of air and the change in pressure was integrated to calculate a tidal volume.

Pain stimulates respiration. Indeed, in several brainstem sites, nociceptive and chemoreceptive functions converge; these include the ventral medulla, 4 the parabrachial complex, 59 and the NTS, 13 areas that all express opioid receptors. It is hypothesized that pain increases tonic input to the respiratory centres, rather than enhancing chemoreflex sensitivity.

The only human laboratory study that has specifically examined the interaction between pain, opioids, and the HCVR 12 demonstrated that pain reversed the respiratory depression induced by morphine, again without affecting the slope of the HCVR. The reversal of opioid-induced respiratory depression by pain can lead to potentially disastrous consequences when alternative analgesic techniques are introduced and highlights the balance between pain and breathing in clinical situations.

This is particularly important, in clinical situations where a patient has received opioids, remains in pain but still breathing , with subsequent neuraxial block causing severe respiratory depression. Genetic studies in knockout mice have given some insight into the mechanisms of action of opioids. Knockout mice lacking the MOP receptor display no analgesia 81 or respiratory depression with morphine, reinforcing clinical observations that analgesic and respiratory effects of opioids are strongly linked.

In humans, interindividual variability in response to opioids may be explained by genetic factors that include sex differences, polymorphisms affecting MOP receptor activity, bioavailability, and metabolism of opioids. In most cases, respiratory and analgesic effects change in parallel, and only one study suggests a potential genetic basis for differential respiratory and analgesic effects. There is an increasing interest in the investigation of sex differences in pain and in the response to analgesics.

Only three, relatively small studies have specifically examined sex differences in opioid-induced respiratory depression 24 , , and each observed a greater respiratory depressant effect in women than in men. In one of these studies, the separate components of the peripheral and central chemoreflex loops were examined. As the strongest effect on respiration was found in the HVR and in the fast peripheral component of the HCVR, the authors hypothesized that these sex differences in the behaviour of the peripheral chemoreflex loop are related to sex steroids.

The phase of menstrual cycle was not controlled in these studies. Given the recent contradictory findings in the analgesic response to opioids, there may be benefit in studying a larger group of subjects.

Polymorphisms of the cytochrome P enzymes CYP2D6 have strong effects on the metabolism of codeine and tramadol. Many studies in humans have compared the respiratory effects of different opioids, including comparisons of drugs with differing potencies, durations of action, partial agonists, and opioids with effects on other receptor systems. Two drugs of particular interest are tramadol and buprenorphine which appear to have differential analgesic and respiratory effects.

This is a synthetic analogue of codeine and a weak MOP opioid receptor agonist. Only one of these studies 89 was performed in conscious humans, where subjective pain assessment was possible.

Unfortunately, the other two studies , are limited by the fact that they were performed in premedicated, anaesthetized patients and that equianalgesia between the two study drugs was not demonstrated. Tramadol-induced respiratory depression has been reported in patients with renal failure.

This is a partial agonist at the MOP receptor, which may cause less respiratory depression than conventional opioids at equianalgesic doses.

Human laboratory studies suggest a ceiling effect in depression of the HCVR, but not in its analgesic effects. It is unclear what specific cellular mechanisms account for the beneficial respiratory profile of buprenorphine. In contrast to the study of brainstem mechanisms where rodent models are used, behavioural and subjective respiratory control mechanisms are ideally suited to being studied in humans, subjects who are able to communicate subjective feelings, and comply with specific tasks.

More recently FMRI has identified, in humans, a network of cortical areas responsible for volitional control of breathing and mediating dyspnoea. Studies of dyspnoea consistently identify the anterior insula, the anterior cingulate, thalamus, and amygdala as brain areas that mediate this unpleasant sensation. Such sensations are ideally suited to FMRI experimental paradigms. Other FMRI studies of conscious respiratory control include one of motor aspects of respiration using a voluntary hyperventilation paradigm 83 and identified activity in cortical and subcortical areas relating to motor control, but not the nociceptive areas identified in the dyspnoea studies.

Currently, there are no imaging studies of opioids and respiration in humans, such studies could dissociate opioid-sensitive and insensitive parts of the respiratory control network. Areas in the brain that are active in response to breath hold. Of these areas the insula, anterior cingulated, and dorsolateral pre-frontal cortex DLPFC express high MOP-receptor binding, and are therefore potential brain regions that may contribute to opioid-induced respiratory depression.

In these images, a colour-coded statistical map of significant activity in eight healthy volunteers is superimposed onto a group mean structural brain scan.

The cross hairs are centred on each signal maxima. Key: a, DLPFC; b, insula; c, putamen; d, cingulate; e, ventrolateral thalamus; f, supramarginal gyrus. V, ventral; R, right. The colour scale indicates the T -score or statistical significance. Figure reproduced with permission from McKay and colleagues. Opioids depress respiration by a number of mechanisms and neuronal sites of action.

It is therefore not surprising that there has been such difficulty in combating opioid-induced respiratory depression. Some of the specific target sites of opioid action on respiratory centres have been elucidated only recently.

The differential effects on rhythm generation and chemoreception suggest that there are many potential therapeutic targets with differing neuronal functions. Both hypoxic and hypercapnic responses are strongly affected by opioids and appear to be strongly mediated in the brainstem. The role of carotid bodies remains unclear, although opioid receptors are expressed here, and they mediate hypoxic and hypercapnic responses; there appears to be a strong case for their transmission being blocked where they input to the NTS in the brainstem.

Enhancing carotid body output may overcome some of these central effects. The effects of opioids on the cortical as opposed to brainstem aspects of respiration have received less attention. Animal models are less suited to such investigations, but in humans, FMRI and PET may provide answers or targets for translation back to animal models. With regard to clinical investigations, there are few studies directly comparing analgesic and respiratory effects in controlled conditions i.

This review of the control of respiration and opioid effects on breathing may provide avenues for further research in humans and in animal models. Google Scholar. Google Preview. Oxford University Press is a department of the University of Oxford. It furthers the University's objective of excellence in research, scholarship, and education by publishing worldwide. Sign In or Create an Account. Sign In. Advanced Search. Search Menu. Article Navigation. Close mobile search navigation Article Navigation.

Volume Article Contents Abstract. Opioid receptors. Brainstem mechanisms of respiration. Measuring opioid effects on breathing in humans. Factors that modulate opioid-induced respiratory depression. Atypical opioids. Cortical effects on breathing. Conclusions and directions for future research. Opioids and the control of respiration. Pattinson K. Nuffield Department of Anaesthetics. Oxford Academic. Cite Cite K. Select Format Select format. Abstract Respiratory depression limits the use of opioid analgesia.

Open in new tab Download slide. Randomised, double blind, placebo controlled crossover trial of sustained release morphine for the management of refractory dyspnoea. Google Scholar Crossref.

Search ADS. Google Scholar PubMed. Morphological and physiological properties of caudal medullary expiratory neurons of the cat. Respiratory depression following oral tramadol in a patient with impaired renal function.