Search

Purpose: A major source of inefficiency in the operating room is the mismatch between scheduled versus actual surgical time. The purpose of this study was to demonstrate a proof-of-concept study for predicting case duration by applying natural language processing (NLP) and machine learning that interpret radiology reports for patients undergoing radius fracture repair. Methods: Logistic regression, random forest, and artificial neural networks (ANN) were tested without NLP and with bag-of-words. Another NLP method tested used ANN and Bidirectional Encoder Representations from Transformers specifically pre-trained on clinical notes (ClinicalBERT). A total of 201 cases were included. The data were split into 70% training and 30% test sets. The average root mean squared error (RMSE) (and 95% confidence interval [CI]) from 10-fold cross-validation on the training set were used to develop each model. Models were then compared to a baseline model, which used historic averages to predict surgical time. Results: The average RMSE was lowest using ANN with ClinicalBERT (25.6 minutes, 95% CI: 21.5 - 29.7), which was significantly (PConclusion: This proof-of-concept study demonstrated the successful application of NLP and machine leaning to extract features from unstructured clinical data resulting in improved prediction accuracy for surgical case duration.

BackgroundThe objective of this study is to develop predictive models for persistent opioid use following lower extremity joint arthroplasty and determine if ensemble learning and an oversampling technique may improve model performance.MethodsWe compared various predictive models to identify at-risk patients for persistent postoperative opioid use using various preoperative, intraoperative, and postoperative data, including surgical procedure, patient demographics/characteristics, past surgical history, opioid use history, comorbidities, lifestyle habits, anesthesia details, and postoperative hospital course. Six classification models were evaluated: logistic regression, random forest classifier, simple-feedforward neural network, balanced random forest classifier, balanced bagging classifier, and support vector classifier. Performance with Synthetic Minority Oversampling Technique (SMOTE) was also evaluated. Repeated stratified k-fold cross-validation was implemented to calculate F1-scores and area under the receiver operating characteristics curve (AUC).ResultsThere were 1042 patients undergoing elective knee or hip arthroplasty in which 242 (23.2%) reported persistent opioid use. Without SMOTE, the logistic regression model has an F1 score of 0.47 and an AUC of 0.79. All ensemble methods performed better, with the balanced bagging classifier having an F1 score of 0.80 and an AUC of 0.94. SMOTE improved performance of all models based on F1 score. Specifically, performance of the balanced bagging classifier improved to an F1 score of 0.84 and an AUC of 0.96. The features with the highest importance in the balanced bagging model were postoperative day 1 opioid use, body mass index, age, preoperative opioid use, prescribed opioids at discharge, and hospital length of stay.ConclusionsEnsemble learning can dramatically improve predictive models for persistent opioid use. Accurate and early identification of high-risk patients can play a role in clinical decision making and early optimization with personalized interventions.

PURPOSE The post-anesthesia care unit (PACU) length of stay is an important perioperative efficiency metric. The aim of this study was to develop machine learning models to predict ambulatory surgery patients at risk for prolonged PACU length of stay - using only pre-operatively identified factors - and then to simulate the effectiveness in reducing the need for after-hours PACU staffing. METHODS Several machine learning classifier models were built to predict prolonged PACU length of stay (defined as PACU stay ≥ 3 hours) on a training set. A case resequencing exercise was then performed on the test set, in which historic cases were re-sequenced based on the predicted risk for prolonged PACU length of stay. The frequency of patients remaining in the PACU after-hours (≥ 7:00 pm) were compared between the simulated operating days versus actual operating room days. RESULTS There were 10,928 ambulatory surgical patients included in the analysis, of which 580 (5.31%) had a PACU length of stay ≥ 3 hours. XGBoost with SMOTE performed the best (AUC = 0.712). The case resequencing exercise utilizing the XGBoost model resulted in an over three-fold improvement in the number of days in which patients would be in the PACU past 7pm as compared with historic performance (41% versus 12%, PCONCLUSION Predictive models using preoperative patient characteristics may allow for optimized case sequencing, which may mitigate the effects of prolonged PACU lengths of stay on after-hours staffing utilization.

Family and twin studies demonstrate that genetic factors determine 20-60% of the vulnerability to opioid use disorder. However, the genes/alleles that mediate the risk of developing addiction-related behaviors, including the sensitivity to the analgesic efficacy of opioids, the development of tolerance, dependence, and escalation of oxycodone taking and seeking, have been ill-defined, thus hindering efforts to design pharmacological interventions to enable precision medicine strategies. Here we characterized oxycodone addiction-like behaviors in heterogeneous stock (HS) rats, that show high genetic diversity that mimics the high genetic variability in humans. HS rats were allowed to self-administer oxycodone for two h/daily for four days (ShA) and then moved to 12h/daily (LgA) for 14 days. Animals were screened for motivation to self-administer oxycodone using a progressive-ratio (PR) schedule of reinforcement and for the development of withdrawal-induced hyperalgesia and tolerance to the analgesic effects of oxycodone using the von-Frey and tail immersion tests, respectively. To reduce cohort-specific effects, we used cohorts of 46-60 rats and normalized the response level within cohorts using a Z-score. To take advantage of the four opioid-related behaviors and further identify subjects that are consistently vulnerable vs. resilient to compulsive oxycodone use, we computed an Addiction Index by averaging normalized responding (Z-scores) for the four behavioral tests. Results showed high individual variability between vulnerable and resilient rats, likely to facilitate the detection of gene variants associated with vulnerable vs. resilient individuals. Such data will have considerable translational value for designing follow-up studies in humans.

RationaleCurrent medications for opioid use disorder include buprenorphine, methadone, and naltrexone. While these medications show significant efficacy in reducing craving and opioid use, there are substantial individual differences in response to these treatments in humans. The reason for such difference is poorly known.ObjectivesHere, we tested the hypothesis that similar individual differences may be observed in a large population of heterogenous stock rats, that have been bred to maximize genetic diversity, using a behavioral paradigm relevant to opioid use disorder.MethodsOver 500 rats were given intermittent (4d/week) and extended access (12h/day) to oxycodone self-administration for 14 sessions to establish oxycodone dependence and escalation of intake. We then measured the effect of buprenorphine (0.5mg/kg), methadone (3mg/kg) and naltrexone (3mg/kg) on the motivation to self-administer oxycodone by using a progressive ratio schedule of reinforcement.ResultsWe found that naltrexone and buprenorphine significantly decreased motivation to oxycodone rewards. While naltrexone reduced oxycodone intake in both males and females, systemic administration with buprenorphine reduced progressive ratio responses only in males. Methadone reduced motivation to oxycodone self-administration in nearly 25% of the population, without reaching statical significance. Our results showed that the efficacy of these medications depends on the severity of addiction like behaviors, indicated by the addiction index.ConclusionsThese results demonstrate individual differences in response to medications to treat opioid use disorder in a genetically diverse population of rats.

The gut brain axis is thought to play a role in behavior and physiological responses through chemical, immunological, and metabolite signaling. Antibiotics, diet, and drugs can alter the transit time of gut contents as well as the makeup of the microbiome. Heterogeneity in genetics and environment are also well-known factors involved in the initiation and perpetuation of substance use disorders. Few viable genetic or biological markers are available to identify individuals who are at risk of escalating opioid intake. Primarily, the addiction field has focused on the nervous system, limiting the discovery of peripheral factors that contribute to addiction. To address this gap, we characterized the microbiome before and after drug exposure, and after antibiotics depletion in male and female heterogenous stock rats to determine if microbiome constituents are protective of escalation. We hypothesized that individuals that are prone to escalation of opioid self-administration will have distinct microbial and metabolic profiles. The fecal microbiome and behavioral responses were measured over several weeks of oxycodone self-administration and after antibiotic treatment. Antibiotic treatment reduces circulating short-chain fatty acids (SCFA) by depleting microbes that ferment fiber into these essential signaling molecules for the gut-brain axis. Depletion of the microbiome increased oxycodone self-administration in a subpopulation of animals (Responders). Supplementation of SCFAs in antibiotic depleted animals decreased elevated oxycodone self-administration. Phylogenetic functional analysis reveals distinct metabolic differences in the subpopulations of animals that are sensitive to antibiotic depletion and animals rescued by SCFA supplementation. In conclusion, this study identifies pre-existing microbiome and metabolic vulnerabilities to escalation of oxycodone self-administration, demonstrates that escalation of oxycodone self-administration dysregulates the microbiome and metabolic landscape, and identifies a causal role of short-chain fatty acids in addiction-like behaviors.

BACKGROUND Expansion of clinical guidance tools is crucial to identify patients at risk of requiring an opioid refill after outpatient surgery. OBJECTIVE The objective of this study was to develop machine learning algorithms incorporating pain and opioid features to predict the need for outpatient opioid refills following ambulatory surgery. METHODS Neural networks, regression, random forest, and a support vector machine were used to evaluate the data set. For each model, oversampling and undersampling techniques were implemented to balance the data set. Hyperparameter tuning based on k-fold cross-validation was performed, and feature importance was ranked based on a Shapley Additive Explanations (SHAP) explainer model. To assess performance, we calculated the average area under the receiver operating characteristics curve (AUC), F1-score, sensitivity, and specificity for each model. RESULTS There were 1333 patients, of whom 144 (10.8%) refilled their opioid prescription within 2 weeks after outpatient surgery. The average AUC calculated from k-fold cross-validation was 0.71 for the neural network model. When the model was validated on the test set, the AUC was 0.75. The features with the highest impact on model output were performance of a regional nerve block, postanesthesia care unit maximum pain score, postanesthesia care unit median pain score, active smoking history, and total perioperative opioid consumption. CONCLUSIONS Applying machine learning algorithms allows providers to better predict outcomes that require specialized health care resources such as transitional pain clinics. This model can aid as a clinical decision support for early identification of at-risk patients who may benefit from transitional pain clinic care perioperatively in ambulatory surgery.

Substance use disorders (SUDs) represent a significant public health crisis. Worldwide, 5.4% of the global disease burden is attributed to SUDs and alcohol use, and many more use psychoactive substances recreationally. Often associated with comorbidities, SUDs result in changes to both brain function and physiological responses. Mounting evidence calls for a precision approach for the treatment and diagnosis of SUDs, and the gut microbiome is emerging as a contributor to such disorders. Over the last few centuries, modern lifestyles, diets, and medical care have altered the health of the microbes that live in and on our bodies; as we develop, our diets and lifestyle dictate which microbes flourish and which microbes vanish. An increase in antibiotic treatments, with many antibiotic interventions occurring early in life during the microbiome's normal development, transforms developing microbial communities. Links have been made between the microbiome and SUDs, and the microbiome and conditions that are often comorbid with SUDs such as anxiety, depression, pain, and stress. A better understanding of the mechanisms influencing behavioral changes and drug use is critical in developing novel treatments for SUDSs. Targeting the microbiome as a therapeutic and diagnostic tool is a promising avenue of exploration. This review will provide an overview of the role of the gut-brain axis in a wide range of SUDs, discuss host and microbe pathways that mediate changes in the brain’s response to drugs, and the microbes and related metabolites that impact behavior and health within the gut-brain axis. Graphic Abstract

BACKGROUND Estimating surgical case duration accurately is an important operating room efficiency metric. OBJECTIVE The primary objective of this 4-year, single academic center retrospective study was to utilize an ensemble learning approach to improve the accuracy of scheduled case duration for spine surgery. The primary outcome measure was case duration. METHODS We compared machine learning models using surgical and patient features to our institutional method, which used historic averages and surgeon adjustment as needed. We implemented multivariable linear regression, random forest, bagging, and XGBoost and calculated average R2, root mean squared error (RMSE), and mean absolute error (MAE) using stratified k-folds cross-validation. We then used the Shapley Additive exPlanations (SHAP) explainer model to determine feature importance. RESULTS 3,315 patients who underwent spine surgery were included. The institution’s current method of predicting case times had poor coefficient of determination with actual times (R2 = 0.19). On k-folds cross-validation, the linear regression model had an R2 of 0.34, RMSE of 165.3, and MAE of 128.4. Among all models, the XGBoost regressor performed the best with an R2 of 0.70, RMSE of 110.9, and MAE of 75.8. Based on SHAP analysis of the XGBoost regression, body mass index, spinal fusions, surgical procedure, and number of spine levels involved were the features with the most impact on the model. CONCLUSIONS Utilizing ensemble learning-based predictive models, specifically XGBoost regression, can improve accuracy of the estimation of spine surgery times. CLINICALTRIAL N/A

Days before surgery, add-ons may be scheduled to fill unused surgical block time at an outpatient surgery center. At times, outpatient surgery centers have time limitations for end of block time and discharge from the postanesthesia care unit (PACU). The objective of our study was to develop machine learning models that predicted the following composite outcome: (1) surgery finished by end of operating room block time and (2) patient was discharged by end of recovery room nursing shift. We compared various machine learning models to logistic regression. By evaluating various performance metrics, including F1 scores, we hypothesized that models using ensemble learning will be superior to logistic regression.Data were collected from patients at an ambulatory surgery center. The primary outcome measurement was determined to have a value of 1 (versus 0) if they met both criteria: (1) surgery ends by 5 pm and (2) patient is discharged from the recovery room by 7 pm. We developed models to determine if a procedure would meet both criteria if it were scheduled at 1 pm, 2 pm, 3 pm, or 4 pm. We implemented regression, random forest, balanced random forest, balanced bagging, neural network, and support vector classifier, and included the following features: surgery, surgeon, service line, American Society of Anesthesiologists score, age, sex, weight, and scheduled case duration. We evaluated model performance with Synthetic Minority Oversampling Technique (SMOTE). We compared the following performance metrics: F1 score, area under the receiver operating characteristic curve (AUC), specificity, sensitivity, precision, recall, and Matthews correlation coefficient.Among 13,447 surgical procedures, the median total perioperative time (actual case duration and PACU length stay) was 165 minutes. When SMOTE was not used, when predicting whether surgery will end by 5 pm and patient will be discharged by 7 pm, the average F1 scores were best with random forest, balanced bagging, and balanced random forest classifiers. When SMOTE was used, these models had improved F1 scores compared to no SMOTE. The balanced bagging classifier performed best with F1 score of 0.78, 0.80, 0.82, and 0.82 when predicting our outcome if cases were to start at 1 pm, 2 pm, 3 pm, or 4 pm, respectively.We demonstrated improvement in predicting the outcome at a range of start times when using ensemble learning versus regression techniques. Machine learning may be adapted by operating room management to allow for a better determination whether an add-on case at an outpatient surgery center could be appropriately booked.

A major limitation of the most widely used current animal models of alcohol dependence is that they use forced exposure to ethanol including ethanol-containing liquid diet and chronic intermittent ethanol (CIE) vapor to produce clinically relevant blood alcohol levels (BAL) and addiction-like behaviors. We recently developed a novel animal model of voluntary induction of alcohol dependence using ethanol vapor self-administration (EVSA). In this model, naive outbred rats given intermittent access to alcohol vapor self-administration exhibit BAL in the 150-300 mg% range and develop somatic signs of withdrawal during acute abstinence. However, it is unknown whether EVSA leads to an escalation of alcohol drinking per se, and whether such escalation is associated with neuroadaptations in brain regions related to stress, reward, and habit. To address these issues, we compared the levels of alcohol drinking during withdrawal between rats passively exposed to alcohol (CIE) or voluntarily exposed to EVSA and measured the number of Fos+ neurons during acute withdrawal (16 h) in the central nucleus of the amygdala (CeA), dorsomedial striatum (DMS), dorsolateral striatum (DLS), nucleus accumbens core (Nacc), periaqueducal grey area (PAG), lateral Habenula (HbL), and the paraventricular nucleus of the thalamus (PVT). The rats were first trained to orally self-administer alcohol in standard operant chambers and then divided in 4 groups (CIE, CI-Air, EVSA and Air-SA) and exposed to intermittent ethanol vapor (passive or active) or intermittent air (passive or active) for 8 h/day, 3 days a week. CIE and EVSA rats exhibited similar BAL (150-300 mg% range) and similar escalation of alcohol drinking during withdrawal, while no changes in terms of drinking were observed in the air exposed rats. CIE and EVSA also increased the motivation for alcohol compared to their respective air control groups under a progressive ratio schedule of reinforcement. Acute withdrawal from EVSA and CIE recruited a similar number of Fos+ neurons in the CeA, however, acute withdrawal from EVSA recruited a higher number of Fos+ neurons in every other brain region analyzed compared to acute withdrawal from CIE. Moreover, acute withdrawal from EVSA specifically recruited the DMS and PVT, a pattern not observed in CIE rats.In summary, these results demonstrate that EVSA produces similar escalation of alcohol drinking, motivation to drink, and blood-alcohol levels than the CIE model, while letting animals voluntary initiate alcohol exposure and maintain alcohol dependence. Moreover, while the behavioral measures of alcohol dependence between the voluntary (EVSA) and passive (CIE) model was similar, the recruitment of neuronal ensembles during acute withdrawal was very different with a higher recruitment of Fos+ neurons in key brain regions important for stress, reward and habit-related processes. The EVSA model may be particularly useful to unveil the neuronal networks and pharmacology responsible for the voluntary induction and maintenance of alcohol dependence and may improve translational studies by providing preclinical researchers with an animal model with better face validity for alcohol use disorder.

Numerous brain regions have been identified as contributing to withdrawal behaviors, but it is unclear the way in which these brain regions as a whole lead to withdrawal. The search for a final common brain pathway that is involved in withdrawal remains elusive. To address this question, we implanted osmotic minipumps containing either saline, nicotine (24 mg/kg/d), cocaine (60 mg/kg/d), or methamphetamine (4 mg/kg/d) for one week in male C57BL/6J mice. After one week, the minipumps were removed and brains collected 8 h (saline, nicotine, and cocaine) or 12 h (methamphetamine) after removal. We then performed single-cell whole-brain imaging of neural activity during the withdrawal period when brains were collected. We used hierarchical clustering and graph theory to identify similarities and differences in brain functional architecture. Although methamphetamine and cocaine shared some network similarities, the main common neuroadaptation between these psychostimulant drugs was a dramatic decrease in modularity, with a shift from a cortical-driven to subcortical-driven network, including a decrease in total hub brain regions. These results demonstrate that psychostimulant withdrawal produces the drug-dependent remodeling of functional architecture of the brain and suggest that the decreased modularity of brain functional networks and not a specific set of brain regions may represent the final common pathway associated with withdrawal.

An emerging element in psychiatry is the gut-brain-axis, the bi-directional communication pathways between the gut microbiome and the brain. A prominent hypothesis, mostly based on preclinical studies, is that individual differences in the gut microbiome composition and drug-induced dysbiosis may be associated with vulnerability to psychiatric disorders including substance use disorder. However, most studies used small sample size, ignored individual differences, or used animal models with limited relevance to addiction. Here, we test the hypothesis that pre-existing microbiome composition and drug-induced changes in microbiome composition can predict addiction-like behaviors using an advanced animal model of extended access to cocaine self-administration in a large cohort of heterogenous stock (HS) rats. Adult male and female HS rats were allowed to self-administer cocaine under short (2h/day) and long access (6h/day) for ~7 weeks under various schedule of reinforcement to identify individuals that are resistant or vulnerable to addiction-like behaviors and fecal samples were collected before the first session and after the last session to assess differences in the microbiome composition. Linear discriminant analysis (LDA) identified sex-dependent and sex-independent differences at the phylum, order, and species level that are differentially abundant in resistant vs. vulnerable individuals, including high level of actinobacteria both before the first exposure to cocaine and after 7 weeks of cocaine self-administration in resistant animals. Predictions of functional gene content using PICRUSt revealed differential regulation of short-chain fatty acid processing in the vulnerable group after self-administration. These results identify microbiome constituents as well as metabolic pathways that are associated with resistance or vulnerability to addiction-like behaviors in rats. Identification of microbes and tangential metabolic pathways involved in cocaine resilience/vulnerability may represent an innovative strategy for the development of novel biomarkers and medication for the treatment of cocaine use disorder.

In addition to its pleasurable effects, weight control is a significant contributor to initiation, maintenance and relapse of cocaine use. This suggests that individual differences in bodyweight control and feeding hormones, such as leptin may contribute to the vulnerability to cocaine use disorder. While pre-clinical studies have shown a mutually inhibitory relationship between leptin and cocaine, they have used small sample sizes and did not investigate individual differences in a large heterogeneous population. Here, we tested if individual differences in bodyweight and blood leptin level is associated with high or low vulnerability to addiction-like behaviors using data from 500 heterogenous stock rats and 160 blood samples from the Cocaine Biobank, using a model of extended access to intravenous self-administration of cocaine. Finally, we tested a separate cohort to evaluate the causal effect of exogenous leptin administration on cocaine seeking. Bodyweight, while changing due to cocaine self-administration in males, was not related to the vulnerability to addiction-like behavior. Blood leptin levels after ~6 weeks of cocaine self-administration did not correlate with addiction-like behaviors, however, baseline blood leptin levels before any access to cocaine negatively predicted addiction-like behavior. Finally, administration of leptin reduced cocaine intake after acute withdrawal and cocaine seeking after 6 weeks of protracted abstinence. These results demonstrate that high blood leptin level before access to cocaine may be a protective factor against the development of cocaine addiction-like behavior, that exogenous leptin reduces the motivation to take and seek cocaine, but that blood leptin level and bodyweight changes in current users are not good biomarkers for addiction-like behaviors.

The COVID-19 pandemic resulted from global infection by the SARS-CoV-2 coronavirus and rapidly emerged as an urgent health issue requiring effective treatments. To initiate infection, the Spike protein of SARS-CoV-2 requires proteolytic processing mediated by host proteases. Among the host proteases proposed to carry out this activation is the cysteine protease cathepsin L. Inhibiting cathepsin L has been proposed as a therapeutic strategy for treating COVID-19. SLV213 (K777) is an orally administered small molecule protease inhibitor that exhibits in vitro activity against a range of viruses, including SARS-CoV-2. To confirm efficacy in vivo, K777 was evaluated in an African green monkey (AGM) model of COVID-19. A pilot experiment was designed to test K777 in a prophylactic setting, animals were pre-treated with 100mg/kg K777 (N=4) or vehicle (N=2) before inoculation with SARS-CoV-2. Initial data demonstrated that K777 treatment reduced pulmonary pathology compared to vehicle-treated animals. A second study was designed to test activity in a therapeutic setting, with K777 treatment (33 mg/kg or 100 mg/kg) initiated 8 hours after exposure to the virus. In both experiments, animals received K777 daily via oral gavage for 7 days. Vehicle-treated animals exhibited higher lung weights, pleuritis, and diffuse alveolar damage. In contrast, lung pathology was reduced in K777-treated monkeys, and histopathological analyses confirmed the lack of diffuse alveolar damage. Antiviral effects were further demonstrated by quantitative reductions in viral load of samples collected from upper and lower airways. These preclinical data support the potential for early SLV213 treatment in COVID-19 patients to prevent severe lung pathology and disease progression.

The gut-brain axis is a bi-directional communication system through which microbial communities in the gut interact with the nervous system, perhaps influencing neuropsychiatric disorders such as drug abuse. This study used behavioral data and biological samples from the Cocaine Biobank to test the hypothesis that the gut microbiota can predict and reflect susceptibility to cocaine reinforcement. Adult male heterogenous rats were catheterized and allowed to self-administer cocaine in short-access sessions (2 hr/day, 10 days, 0.5 mg/kg per intravenous infusion), followed by progressive ratio (PR) tests, long-access sessions (6 hr/day, 14 days), and alternating blocks of PR, long-access, and footshock testing. Fecal samples were collected at three time points and bacterial 16s rRNA genes were sequenced to profile the microbiota. As expected, cocaine-related behavior varied among subjects, such that a quartile split identified low and high responders on each measure, as well as an overall addiction index. Although beta diversity in the microbiota at baseline and after short access did not predict membership in high or low addiction quartiles, linear discriminant analysis (LDA) identified taxa that were more robustly represented in low or high responders. Beta diversity after long access was different among quartiles, as were several specific taxa. Investigation of baseline differences revealed that high relative abundance of Akkermansia muciniphila predicted future low response rates, whereas Ruminococcaceae predicted high response. This study is the first to report that microbiota variability reflects levels of cocaine intake and that microbial profiles might facilitate diagnosis and identify risk factors predictive of drug vulnerability.Significance StatementMicrobial organisms inhabiting the gut of animals appear to influence organismal function through various signaling pathways, ultimately affecting behavior and disease vulnerability. This experiment investigates links between gut bacteria and vulnerability to addiction-related behaviors in adult male rats. Not only did gut bacterial profiles change as a result of cocaine intake but also gut bacterial profiles before any exposure to cocaine predicted which animals would be high or low addiction-prone individuals. These results suggest that microbial profiles might facilitate diagnosis and identify risk factors predictive of drug addiction.

Visual Abstract, The rat oxycodone and cocaine biobanks contain samples that vary by genotypes (by using genetically diverse genotyped HS rats), phenotypes (by measuring addiction-like behaviors in an advanced SA model), timepoints (samples are collected longitudinally before, during, and after SA, and terminally at three different timepoints in the addiction cycle: intoxication, withdrawal, and abstinence or without exposure to drugs through age-matched naive rats), samples collected (organs, cells, biofluids, feces), preservation (paraformaldehyde-fixed, snap-frozen, or cryopreserved) and application (proteomics, transcriptomics, microbiomics, metabolomics, epigenetics, anatomy, circuitry analysis, biomarker discovery, etc. Substance use disorders (SUDs) are pervasive in our society and have substantial personal and socioeconomical costs. A critical hurdle in identifying biomarkers and novel targets for medication development is the lack of resources for obtaining biological samples with a detailed behavioral characterization of SUD. Moreover, it is nearly impossible to find longitudinal samples. As part of two ongoing large-scale behavioral genetic studies in heterogeneous stock (HS) rats, we have created two preclinical biobanks using well-validated long access (LgA) models of intravenous cocaine and oxycodone self-administration (SA) and comprehensive characterization of addiction-related behaviors. The genetic diversity in HS rats mimics diversity in the human population and includes individuals that are vulnerable or resilient to compulsive-like responding for cocaine or oxycodone. Longitudinal samples are collected throughout the experiment, before exposure to the drug, during intoxication, acute withdrawal, and protracted abstinence, and include naive, age-matched controls. Samples include, but are not limited to, blood plasma, feces and urine, whole brains, brain slices and punches, kidney, liver, spleen, ovary, testis, and adrenal glands. Three preservation methods (fixed in formaldehyde, snap-frozen, or cryopreserved) are used to facilitate diverse downstream applications such as proteomics, metabolomics, transcriptomics, epigenomics, microbiomics, neuroanatomy, biomarker discovery, and other cellular and molecular approaches. To date, >20,000 samples have been collected from over 1000 unique animals and made available free of charge to non-profit institutions through https://www.cocainebiobank.org/ and https://www.oxycodonebiobank.org/.

The two most prevalent substance use disorders involve alcohol (alcohol use disorder) and nicotine (tobacco use disorder), which are often co-abused. Robust preclinical and translational evidence indicates that individuals initiate drug use for the acute rewarding effects of the substance. The development of negative emotional states is key for the transition from recreational use to substance use disorders as subjects seek the substance to obtain relief from the negative emotional states of acute withdrawal and protracted abstinence. The neuropeptide corticotropin-releasing factor (CRF) is a major regulator of the brain stress system and key in the development of negative affective states. The present review examines the role of CRF in preclinical models of alcohol and nicotine abuse and explores links between CRF and anxiety-like, dysphoria-like, and other negative affective states. Finally, the present review discusses preclinical models of nicotine and alcohol use with regard to the CRF system, advances in molecular and genetic manipulations of CRF, and the importance of examining both males and females in this field of research.

RATIONALE: Over the last decade, oxycodone has become one of the most widely abused drugs in the USA. Oxycodone use disorder (OUD) is a serious health problem that has prompted a need to develop animal models of OUD that have both face and predictive validity. Oxycodone use in humans is more prevalent in women and leads to pronounced hyperalgesia and irritability during withdrawal. However, unclear is whether current animal models of oxycodone self-administration recapitulate these characteristics in humans. OBJECTIVES: We assessed the face validity of a model of extended-access oxycodone self-administration in rats by examining the escalation of oxycodone intake and behavioral symptoms of withdrawal, including irritability-like behavior and mechanical nociception, in male and female Wistar rats. RESULTS: Both male and female rats escalated their oxycodone intake over fourteen 12-h self-administration sessions. After escalation, female rats administered more drug than male rats. No differences in plasma oxycodone levels were identified, but males had a significantly higher level of oxycodone in the brain at 30 min. Extended access to oxycodone significantly decreased aggressive-like behavior and increased defensive-like behaviors when tested immediately after a 12-h self-administration session, followed by a rebound increase in aggressive-like behavior 12 h into withdrawal. Tests of mechanical nociception thresholds during withdrawal indicated pronounced hyperalgesia. No sex differences in irritability-like behavior or pain sensitivity were observed. CONCLUSIONS: The present study demonstrated the face validity of the extended access model of oxycodone self-administration by identifying sex differences in the escalation of oxycodone intake and pronounced changes in pain and affective states.

Numerous brain regions have been identified as contributing to addiction-like behaviors, but unclear is the way in which these brain regions as a whole lead to addiction. The search for a final common brain pathway that is involved in addiction remains elusive. To address this question, we used male C57BL/6J mice and performed single-cell whole-brain imaging of neural activity during withdrawal from cocaine, methamphetamine, and nicotine. We used hierarchical clustering and graph theory to identify similarities and differences in brain functional architecture. Although methamphetamine and cocaine shared some network similarities, the main common neuroadaptation between these psychostimulant drugs was a dramatic decrease in modularity, with a shift from a cortical- to subcortical-driven network, including a decrease in total hub brain regions. These results demonstrate that psychostimulant withdrawal produces the drug-dependent remodeling of functional architecture of the brain and suggest that the decreased modularity of brain functional networks and not a specific set of brain regions may represent the final common pathway that leads to addiction.Significance StatementA key aspect of treating drug abuse is understanding similarities and differences of how drugs of abuse affect the brain. In the present study we examined how the brain is altered during withdrawal from psychostimulants. We found that each drug produced a unique pattern of activity in the brain, but that brains in withdrawal from cocaine and methamphetamine shared similar features. Interestingly, we found the major common link between withdrawal from all psychostimulants, when compared to controls, was a shift in the broad organization of the brain in the form of reduced modularity. Reduced modularity has been shown in several brain disorders, including traumatic brain injury, and dementia, and may be the common link between drugs of abuse.

Substance use disorders have a complex etiology. Genetics, the environment, and behavior all play a role in the initiation, escalation, and relapse of drug use. Recently, opioid use disorder has become a national health crisis. One aspect of opioid addiction that has yet to be fully examined is the effects of alterations of the microbiome and gut-brain axis signaling on central nervous system activity during opioid intoxication and withdrawal. The effect of microbiome depletion on the activation of neuronal ensembles was measured by detecting Fos-positive (Fos+) neuron activation during intoxication and withdrawal using a rat model of oxycodone dependence. Daily oxycodone administration (2 mg/kg) increased pain thresholds and increased Fos+ neurons in the basolateral amygdala (BLA) during intoxication, with a decrease in pain thresholds and increase in Fos+ neurons in the periaqueductal gray (PAG), central nucleus of the amygdala (CeA), locus coeruleus (LC), paraventricular nucleus of the thalamus (PVT), agranular insular cortex (AI), bed nucleus of the stria terminalis (BNST), and lateral habenula medial parvocellular region during withdrawal. Microbiome depletion produced widespread but region- and state-specific changes in neuronal ensemble activation. Oxycodone intoxication and withdrawal also increased functional connectivity among brain regions. Microbiome depletion resulted in a decorrelation of this functional network. These data indicate that microbiome depletion by antibiotics produces widespread changes in the recruitment of neuronal ensembles that are activated by oxycodone intoxication and withdrawal, suggesting that the gut microbiome may play a role in opioid use and dependence. Future studies are needed to better understand the molecular, neurobiological, and behavioral effects of microbiome depletion on addiction-like behaviors.

The activation of a neuronal ensemble in the central nucleus of the amygdala (CeA) during alcohol withdrawal has been hypothesized to induce high levels of alcohol drinking in dependent rats. In the present study we describe that the CeA neuronal ensemble that is activated by withdrawal from chronic alcohol exposure contains ~80% corticotropin-releasing factor (CRF) neurons and that the optogenetic inactivation of these CeA CRF+ neurons prevents recruitment of the neuronal ensemble, decreases the escalation of alcohol drinking, and decreases the intensity of somatic signs of withdrawal. Optogenetic dissection of the downstream neuronal pathways demonstrates that the reversal of addiction-like behaviors is observed after the inhibition of CeA CRF projections to the bed nucleus of the stria terminalis (BNST) and that inhibition of the CRFCeA-BNST pathway is mediated by inhibition of the CRF-CRF1 system and inhibition of BNST cell firing. These results suggest that the CRFCeA-BNST pathway could be targeted for the treatment of excessive drinking in alcohol use disorder., Withdrawal from alcohol activates neurons in the central amygdala (CeA) and increases craving for alcohol. The authors show that these neurons predominantly express CRF and project to the BNST. Inactivation of this pathway reduces the dependence-related escalation of alcohol drinking.

Voluntary urination ensures that waste is eliminated when safe and socially appropriate, even without a pressing urge. Uncontrolled urination, or incontinence, is a common problem with few treatment options. Normal urine release requires a small region in the brainstem known as Barrington’s nucleus (Bar), but specific neurons that relax the urethral sphincter and enable urine flow are unknown. Here we identify a small subset of novel Bar neurons that control the urethral sphincter in mice. These excitatory neurons express estrogen receptor 1 (BarESR1), project to sphincter-relaxing interneurons in the spinal cord and are active during natural urination. Optogenetic stimulation of BarESR1 neurons rapidly initiates sphincter bursting and efficient voiding in anesthetized and behaving animals. Conversely, optogenetic and chemogenetic inhibition reveals their necessity in motivated urination behavior. The identification of these cells provides an expanded model for the control of urination and its dysfunction.

Urination may occur either reflexively in response to a full bladder or deliberately irrespective of immediate need. Voluntary control is desired because it ensures that waste is expelled when consciously desired and socially appropriate1,2. Urine release requires two primary components: bladder pressure and urethral relaxation1–3. Although the bladder contracts during urination, its slow smooth muscle is not under direct voluntary control and its contraction alone is not sufficient for voiding. The decisive action of urination is at the urethral sphincter, where striated muscle permits fast control. This sphincter is normally constricted, but relaxes to enable urine flow. Barrington’s nucleus (Bar, or pontine micturition center) in the brainstem is known to be essential for the switch from urine storage to elimination4–7, and a subset of Bar neurons expressing corticotropin releasing hormone (BarCRH) have recently been shown to promote bladder contraction8–10. However, Bar neurons that relax the urethral sphincter to enable urination behavior have not been identified. Here we describe novel brainstem neurons that control the external urethral sphincter. We find that scent marking behavior in male mice depends upon a subpopulation of spatially clustered Bar neurons that express high levels of estrogen receptor 1 (BarESR1). These neurons are glutamatergic, project to urinary nuclei in the spinal cord with a bias towards sphincter-inhibiting interneurons, and their activity correlates with natural urination. Optogenetic stimulation of BarESR1 neurons rapidly initiates sphincter bursting and efficient voiding in absence of sensory cues in anesthetized and behaving animals. Conversely, inhibiting the activity of these neurons prevents olfactory cues from promoting scent marking behavior. The identification of BarESR1 cells provides an expanded model for the supraspinal control of urination and its dysfunction.

CLC chloride channels and transporters are vital to many physiological processes. Defects in these proteins can lead to diseases that affect muscle, cardiovascular system, kidneys, and bones. Historically, high affinity inhibitors have played crucial roles in the characterization of ion transporter structure and function and in the treatment of disease. However, there is a dearth of known CLC inhibitors. To search for improved inhibitors, we synthesized derivatives of DIDS (4,4′-diisothio-cyanostilbene-2,2′-disulfonic acid), a small molecule that inhibits several CLC proteins with low affinity and specificity. We found that linking DIDS and octanoic acid generated a CLC-ec1 inhibitor (OADS) that inhibits CLC-ec1 with low micromolar affinity. We are characterizing the mechanism of inhibition using flux assays, electrophysiology, isothermal titration calorimetry, photoaffinity labeling, and mutagenesis. Given that OADS and OADS derivatives that inhibit ClC-ec1 are rather hydrophobic, we also explored their potential to alter lipid bilayer properties using a gramicidin based fluorescence assay.