Acute Comparative Analysis of Anesthetic Methods in Neonatal Mice (Mus musculus) Using Ultrasonic Vocalizations and Biochemical Markers

Abstract

Hypothermia has been used as an anesthetic technique in neonatal rodents since 1934. While the long-term impacts of neonatal anesthetic exposure in adult mice are still under investigation, its acute effect during the neonatal period is poorly characterized. Although hypothermia can achieve sufficient anesthetic depth to eliminate pain sensation, its effect during induction and recovery remains unclear. Mice emit ultrasonic vocalizations (USVs) within a range of 30-120 kHz for normal behavior and communication, with lower frequencies often associated with pain and distress. Using USV analysis and other physiologic measures, we assessed the well-being of neonatal mice subjected to different anesthetic exposures at different ages (postnatal day [PND] 2, 5, 8, 11, and 14). C57BL/6 neonatal mice (n = 483) were randomly assigned to 1 of 7 groups: no handling or anesthesia, handling only, hypothermia (PND 2, 5, and 8 only), isoflurane vaporizer, sevoflurane vaporizer, isoflurane drop, or sevoflurane drop methods. USVs were recorded at 4 time points: before the anesthetic exposure (T0), immediately during recovery (DR), 5 minutes after recovery (PR), and 120 minutes after recovery (T120) (n ≥ 12 pups per group). Induction and recovery were timed. At T120, terminal blood was collected to measure stress-related biochemical markers (blood glucose, corticosterone, norepinephrine). Our results show that neonatal mice receiving hypothermia anesthesia had significantly lower frequency of USVs during and after recovery, longer duration of USV calls, and longer recovery time compared with other anesthetic groups. Our findings suggest that hypothermia anesthesia induces acute distress in neonatal mice ≤PND 8 compared with inhalational anesthetics. In accordance with the principles of the 3Rs to minimize pain and distress, using inhalant anesthesia (isoflurane or sevoflurane) is preferable to hypothermia for neonatal anesthesia in mouse pups at ≤PND 8.

Introduction

Neonatal mice have increasingly been used in neurologic, behavioral, and cardiovascular studies.^1–4 They have become a model for complex diseases such as hypoxic ischemia encephalopathy, juvenile spinal cord injury, and intraventricular hemorrhage.^5–7 For many studies anesthesia is frequently required, especially for procedures demanding a high degree of precision and accuracy. However, neonatal anesthesia presents unique challenges. Due to their small size and altricial status, neonatal mice are at particular risk of overdose from injectable anesthesia even at small volumes. Although injectable anesthesia drugs such as ketamine, xylazine, avertin, and propofol have been successfully used in pups 7 days and older,^8,9 hypothermia and inhalant anesthesia remain the most commonly used methods for neonatal anesthesia starting at birth, or postnatal day (PND) 0.

For decades, the most common anesthetic method in neonates <10 days old has been hypothermia.¹⁰ This method requires inexpensive materials (water and ice) and a pups’ poikilothermic status to perform invasive procedures during the neonatal period.^9,11 Although neonates are highly resistant to brain damage at extremely low temperatures, returning their bodies to normal body temperature has been associated with rewarming complications such as thermal injuries.¹¹ Inhalant anesthesia does not rely on a drastic temperature drop to induce a depth of anesthesia. Instead, use of inhalant anesthetics such as isoflurane and sevoflurane have been associated with more normal physiologic parameters in rodents, such as heart rate, reparatory rate, and oxygen levels, when compared with hypothermia.^12,13 Unlike hypothermia,^14,15 neonatal inhalant anesthesia events have been shown to induce cognitive impairment, neurotoxicity, and behavior changes in adult mice.^16–18

In accordance with the 3Rs principles (replacement, reduction, refinement),¹⁹ we sought to further refine neonatal anesthesia practices to minimize distress, pain, and suffering of animals. To date, no published studies have evaluated pain among hypothermia and anesthesia methods in neonatal mice. Given that hypothermia causes painful skin burns in humans,²⁰ and the Guide for the Care and Use of Laboratory Animals states “unless the contrary is known or established, it should be considered that procedures that cause pain in humans may also cause pain in other animals,”²¹ evaluating these techniques is both scientifically and ethically significant.

Traditionally behavioral assessments such as tail flick, hot plate, von Frey filaments, or other clinical parameters (facial grimace scale, cage side observations) are used in adult mice.^19,22 Since neonatal mice are small in size, are continuing to develop facial features and the nervous system, and lack body coordination, these tests are unlikely to yield reliable results. Instead, ultrasonic vocalizations (USVs) can be recorded and used to characterize rodent calls as aversive or appetitive based on frequency, duration, and call structure.²³ Adult rodent USVs have been associated with different behaviors such as exploring, alerting, and reproductive related calls,^24–26 whereas rodent neonates’ USVs are often only associated with isolation, and characterized as distress calls warranting dam retrieval.^23,27–29 Because rodent neonates are altricial, any change to their homeostasis can signal distress and naturally require dam retrieval,^27,30 supporting USVs as a reliable sign of distress in neonatal experimental manipulations. In addition, blood levels of glucose, corticosterone, and norepinephrine are increasingly used as indirect pain indicators.²² These parameters reflect the animal’s stress response and help assess distress or pain in neonatal mice.^3,15,31,32

Our study assessed the well-being of neonatal C57BL/6 mice subjected to hypothermia and different inhalant anesthetic exposures (AEs) at different postnatal ages (PND 2, 5, 8, 11, and 14). We used USV, blood glucose, corticosterone, and norepinephrine analysis as stress indicators. C57BL/6 mice were used due to their wide use in translational studies, including but not limited to behavioral, USV, and phenotyping studies.¹⁹ We hypothesized that neonatal mice undergoing hypothermia anesthesia would experience increased distress levels compared with those undergoing inhalant anesthesia.

Materials and Methods

Ethical review.

All experiments were approved by the University of North Carolina at Chapel Hill IACUC. All mice were treated in accordance with the Guide for the Care and Use of Laboratory Animals in a facility accredited by AAALAC International.

Experimental animals.

A total of 483 C57BL/6 neonatal mice (n = 236 females, 247 males) were used in this study. Mice were bred in-house using harem breeding of C57BL/6J donated breeders. Once pregnancy was confirmed, female breeders were separated into individual cages. One litter was used per female breeder; the average litter size was 6 pups, with a total of 87 litters. All pups were included in this study if they had milk spots at the time of experiment. Animals were free of pinworms (Syphacia muris, Syphacia obvelata, Aspiculuris tetraptera), fur mites (Myocoptes, Radfordia, Myobia), epizootic diarrhea of infant mice, mouse hepatitis virus, mouse parvovirus, minute virus of mice, Theiler’s mouse encephalomyelitis virus/GDVII, ectromelia virus, lymphocytic choriomeningitis virus, mouse adenovirus-1/2, murine cytomegalovirus, polyoma virus, pneumonia virus of mice, reovirus 3, Sendai virus, Mycoplasma pulmonis, and Filobacterium rodentium, as monitored by soiled-bedding sentinels and quarterly exhaust air dust testing.

Animal housing conditions.

Mice were housed in individually ventilated cages (7 × 10 × 7 cm; GM500; Tecniplast, Buguggiate, Italy) on corncob bedding (¼ in., irradiated, Bed-O-Cobs; Lab Supply, Durham, NC) with red rodent forts (K3588, red, certified, mouse fort; Bio-Serv, Flemington, NJ), a cotton square for nesting material (Ancare, Bellmore, NY), and autoclaved crinkle paper as enrichment and nesting material. The room had a 12-hour light/12-hour dark cycle (lights on at 7:00 am; lights off at 7:00 pm) at 70-74 °F (21.1-23.3 °C) and 30%-70% relative humidity. Mice were provided with food (5V5R PicoLab select rodent 50 IF/6F; LabDiet, Richmond, IN) and reverse osmosis–filtered water ad libitum. Cages were changed every 2 weeks.

Experimental design.

On the day of testing, the dam was transferred from its home cage to a static cage. The home cage containing the litter was placed on a Tatown heating pad. Each pup was individually handled, weighed, identified with a permanent marker, and randomly assigned to an experimental group using Excel to ensure unbiased distribution across groups. Baseline temperature was measured using a rodent-specific infrared thermometer (non-contact infrared thermometer; Kent Scientific, Torrington, CT). Based on pilot study findings, temperatures were recorded dorsally between the shoulders for pups aged 5 days or younger and ventrally along the midline between the hips for pups aged 8 days or older.

Pups assigned to the control group (no handling) were decapitated for blood collection within 5 minutes of being weighed and identified. All other pups were individually placed in USV chambers. Based on pilot study findings, pups aged 8 days or younger were placed in glass containers with one squared layer of paper towel. Pups aged 11 or older were placed in tall plastic containers. Four USV chambers were used at a time. For litters with >4 pups, the remaining pups were returned to the home cage with dam and placed in their nest until USV chambers were sanitized and unoccupied. Containers were sanitized with Peroxigard (Peroxigard, Oakville, ON, Canada), and the paper towel square was replaced between use. When pups were not being individually handled, under anesthesia, or in a USV chamber, they remained with their dam in their home cage.

After baseline data collection, AEs were performed on individual pups according to their assigned AE group (handling only, hypothermia, isoflurane vaporizer, sevoflurane vaporizer, isoflurane drop, or sevoflurane drop). Following AE, each pup’s temperature was recorded, and the pup was placed on its back in a USV chamber with heat support. USVs were recorded during recovery until the pup regained consciousness signaled by a positive righting reflex. For the handling only group, during recovery, recordings lasted 5 minutes. For all other groups, recovery was defined as the pup regaining the righting reflex, confirmed by video monitoring. If a pup did not regain the righting reflex within 5 minutes, it was stimulated with a toe pinch and returned to the USV chamber for additional recovery USV recordings. Immediately after recovery was confirmed, a separate 5-minute USV recording was collected after recovery. Following the postrecovery recording, the pup was returned to its home cage with dam to resume normal activity. At 120 minutes after the pup regained its righting reflex, its temperature was recorded, and it was placed in the USV chamber for the final 5-minute USV recording (T120). Immediately after the T120 recording, the pup was decapitated and blood was collected. Blood was collected in mini microcontainers (BD Microtainer SST, ref. 365967; Becton Dickinson, Franklin Lakes, NJ), centrifuged for at least 5 minutes, and stored in a −80 °C (−112 °C) freezer until further analysis. All experimental procedures were conducted between 7:00 am and 5:00 pm, with blood collection for each litter occurring approximately every 2.5-3 hours, reflecting the relative length of each procedure.

Anesthetic Exposure.

Seven experimental groups were assigned for each postnatal age group: no handling or anesthesia, handling only, hypothermia, isoflurane vaporizer, sevoflurane vaporizer, isoflurane drop, and sevoflurane drop (Table 1). Nine pups died during the study due to anesthetic overdose or failure to recover from anesthesia, including 3 sevoflurane vaporizer-anesthetized pups at PND 11, 2 sevoflurane vaporizer-anesthetized pups at PND 8, 2 isoflurane drop-anesthetized pups at PND 2, and 2 sevoflurane drop-anesthetized pups at PND 2. These pups were excluded from Table 1 and data analysis.

Table 1.

Sample Size Grouping

Group		PND 2	PND 5	PND 8	PND 11	PND 14
No handling	Female	8	7	7	8	6
	Male	8	7	10	9	7
Handling only	Female	7	8	9	7	6
	Male	8	8	9	7	6
Hypothermia	Female	7	7	7	N/A	N/A
	Male	7	7	6	N/A	N/A
Isoflurane vaporizer	Female	7	7	7	7	7
	Male	7	8	7	8	6
Sevoflurane vaporizer	Female	7	7	8	8	7
	Male	8	7	9	6	7
Isoflurane drop	Female	7	7	7	8	6
	Male	8	7	11	6	7
Sevoflurane drop	Female	6	7	6	8	7
	Male	7	7	8	7	8

Time of induction, time to reach deep anesthetic plane, end time of anesthesia, and time of recovery were recorded for all groups. All groups, except for the isoflurane (Isospire; Dechra, Overland Park, KS) and sevoflurane drop groups, were maintained under a deep anesthetic plane for 10 minutes. Isoflurane and sevoflurane drop groups were maintained under a deep anesthetic plane for 2 minutes. Deep anesthetic planes were confirmed by the loss of pedal withdrawal reflex of all 4 limbs, tested by using a consistent, firm pinch to the digits on each paw. All paws were assessed to ensure complete loss of sensation consistent with a deep plane of anesthesia. During the pilot study, the use of hemostats caused trauma to neonatal paws and was therefore discontinued. Instead, all procedures in the full study were performed by the same experienced veterinarian to minimize variability in technique. A camera (USB HD webcam model JVCU100; j5create, Taiwan) was used to monitor time of recovery in USV chambers.

Handling-only pups were placed in a short plastic holding container on a heating pad for 10 minutes in a chemical fume hood, serving as a secondary control in place of an AE. They were not handled for any additional time in relationship to the other anesthetic groups. All AEs were conducted in a chemical fume hood. Hypothermia-anesthetized pups were placed on a layer of nitrile material, and placed in an ice bath, ensuring that they did not directly contact ice or water. Body temperatures were monitored and ranged between 53 and 79 °F (12 and 26 °C) during cryoanesthesia, with younger pups having lower range values. After initial cooling, the pups were positioned on their backs for the remainder of the AE (10 minutes) in room air. The ice bath was replenished every 1-2 uses.

For the isoflurane and sevoflurane (sevoflurane; Dechra, Overland Park, KS) vaporizer anesthesia, pups were positioned in a nose cone attached to a rodent nonrebreathing circuit. The oxygen flow rate was set at 2 L/min at 100% O₂ for both groups. Isoflurane induction was set at 4%-5% and reduced to 2%-3% for maintenance. Sevoflurane induction was set at 5%-6% and reduced to 3%-4% for maintenance. Exact vaporizer setting was adjusted individually based on the pup’s anesthetic depth. Depth was assessed using established criteria for neonatal rodents, including reduced spontaneous movement and the absence of a pedal withdrawal reflex. Because neonatal pups have immature nervous systems and exhibit minimal responses to traditional reflex testing, the pedal withdrawal reflex was used cautiously and interpreted in combination with respiratory rate, muscle tone, and overall activity every 3-4 minutes. Adjustments to anesthetic concentration were made only when necessary to maintain a stable plane of anesthesia while minimizing exposure time.

For isoflurane and sevoflurane drop methods, a 15-mL conical tube was used to hold the liquid inhalant. First, 60 μL of either isoflurane or sevoflurane was placed in the conical tube. Then, one layer of 2 × 2 sponge gauze was placed inside the tube to prevent pups from contacting the liquid inhalant. The conical tube was closed for 1 minute. Then, a pup was placed shoulder-deep into the conical tube. If a deep anesthetic plane was not reached within 5 minutes, an additional 20 μL of either isoflurane or sevoflurane was added to the conical tube.

For all inhalant anesthesia groups, body temperatures were controlled by providing heat support throughout the experimental procedure. All experimental groups were exposed to room air oxygen at all times, excluding inhalant anesthesia groups who had 100% oxygen via nose cone during their anesthesia event.

USV chambers.

Four sound-attenuating chambers were used to record USVs as previously described.^24,31,33 Briefly, each chamber was equipped with a camera (USB HD webcam model JVCU100; j5create, Taiwan) Velcroed 20 cm above the chamber floor (left-aligned), and an ultrasonic microphone (UltraSoundGate CM16/CMPA; Avisoft Bioacoustics, Glienicke, Germany) was suspended 20 cm above the floor (centered). The cords exited through the top of the chamber and were connected to a computer and the recording device (UltraSoundGate 116Hb; Avisoft Bioacoustics. Glienicke, Germany), respectively. The recording software (RECORDER USGH; Avisoft Bioacoustics, Glienicke, Germany) was configured with a sampling rate of 250,000 Hz, fast Fourier transform length of 256 points, time-window overlap of 50%, flat top-weighted window, and 16-bit format. A heating pad (TATOWN), covered with a plastic bag, was placed on the floor of each chamber. Recordings were conducted in a quiet room with a running chemical fume hood. Chambers were sanitized with Peroxigard (Peroxigard, Oakville, ON, Canada) at the start and end of each experimental day.

USV recordings.

For baseline recording, each pup was placed in the testing chamber and allowed to acclimate to the closed chamber for 5 minutes, then a 5-minute baseline recording (T0) was collected. Additional USV recordings were also collected during recovery, 5 minutes after recovery, and 120 minutes after recovery (T120). Pups were not taken out of the USV chamber between during recovery and after recovery recordings.

Quantification of USVs.

For acoustical analysis, recordings collected from pups were transferred to Avisoft SAS Lab Pro (version 5.3.00, Avisoft Bioacoustics) as previously described.^24,31,33 Fast Fourier transformation was used to generate spectrograms with parameters of fast Fourier transform length of 256, temporal resolution overlap of 75%, 100% frame size, and flat top-weighted window. Acoustic signals were detected via the program’s automatic whistle tracking algorithm, with a threshold of −20 dB, a minimum duration of 3 ms, and a hold time of 2 ms. Each USV recording was set to record for 5 minutes. If pups took >5 minutes to recover, each subsequent USV recording was consolidated and counted as one USV recording for data analysis. Frequencies <20 Hz were not included in data analysis to reduce any background noise outside the relevant frequency. Parameters analyzed for each USV recording included number of calls, durations of calls (s), and mean frequencies (kHz).

Blood serum collection.

Blood glucose.

Serum blood glucose was measured by using a glucometer (CareSens N blood glucose monitoring system; i-SENS, Torrance, CA) that underwent daily calibration prior to experimental decapitation.

ELISAs.

Blood serum measurements of corticosterone were made using the DetectX corticosterone enzyme immunoassay kit (K014-H5; Arbor Assays, Ann Arbor, MI). Samples and standards were treated according to the manufacturer’s instructions. Each sample and standard were measured in duplicate. Similar instructions were followed for blood serum measurements of norepinephrine using the noradrenaline (competitive enzyme immunoassay) ELISA kit (LS-F5550; LS Bio, Shirley, MA). Norepinephrine samples were diluted to a 1:30 ratio. Corticosterone samples remained undiluted.

A standard curve was created for every analysis. The plates were read on an ELISA plate reader set (Epoch; BioTek, Winooski, VT) to 450 nm by using Gen5 software (BioTek, version 3.11). Concentrations were calculated by using the 4-parameter logistic curve assay on MyAssays.com. Corticosterone values were returned as ‘high’ when >250,000 ng/mL and were set to 250,000 in the analyses (n = 25 of 483). A standard curve was created for every analysis.

Statistical methods.

PND 11 and PND 14 pups did not have a hypothermia anesthesia group, as hypothermia is not appropriate to use in pups older than 10 days according to the 2020 AVMA euthanasia guidelines.¹⁰ Therefore, data were separated into 2 groups, that is, younger pups and older pups, for analysis. Younger pups included data from PND 2, 5, and 8. Older pups included data from PND 11 and 14. Sample sizes were based on a power analysis that a group size of 14 pups was sufficient to detect a typical effect size and variance, power of 0.80, and an α level of 0.05 for all outcome measures.

To examine the trend in each outcome change trajectory over time, model fitting proceeded as follows for all USV outcome variables (frequency, duration, and calls). The statistician was blinded to USV data. Hierarchical linear mixed models were fitted to each outcome using the R package lme4. Models were fitted via restricted maximum likelihood estimation. To account for the nested nature of the data, 2-level nested effects models with a random intercept for each pup and each unique combination of pup and time were fitted. Time was examined as a continuous variable using linear, quadratic, and cubic functional forms. A quadratic specification of time was chosen based on Akaike information criterion and Bayesian information criterion as well as the ability of these models to converge successfully when including all predictors of interest. Random slopes for time were fitted; however, these models would not converge. Therefore, final models are random intercept only models. The main predictors of interest, that is, sex, PND, time, and group, and interactions of group*time, group*PND and group*time*PND were examined. For parsimony, group*time*PND and group*PND interactions that were not statistically significant were not retained in final models. Residual diagnostics were assessed for all models. Except for the model for total number of calls, all models had acceptable residual diagnostics. The model for total number of calls was subsequently specified as a Poisson mixed model, but this model failed to converge. For all models that converged successfully, a series of planned contrasts for post hoc analyses were conducted using the Tukey honestly significant difference test to adjust for multiple comparisons. Specifically, changes across groups over time and changes over time within groups were compared. For all post hoc analyses, covariates were held at their mean. All graphed USV data are shown as predicted values, representing the model-implied means after adjusting for relevant covariates.

Assumptions of normal distribution were not met based on visual assessment of quantile-quantile plots for all non-USV data. Nonparametric comparison tests, that is, Kruskal-Wallis and Wilcoxon tests, were applied to the following data: time of induction, time of recovery, blood glucose, corticosterone, and norepinephrine serum by age group, older pups (PND 11 and 14) and younger pups (PND 2, 5, and 8). Values of P < 0.05 were considered significant. All non-USV tests were performed using SAS JMP Pro 17 and in collaboration with the Odom Institution at the University of North Carolina-Chapel Hill.

Results

The following results are presented based on individual comparisons among the group. No single control was used within the anesthetic groups, as there is currently no consensus in the scientific field regarding which neonatal anesthetic regimen is superior.

USVs.

Table 2 summarizes the mean frequency, duration, and total number of USVs during this study.

Table 2.

Descriptive Statistics for USV Data

Outcomea	Hypothermia	Handling only	Isoflurane drop	Isoflurane vaporizer	Sevoflurane drop	Sevoflurane vaporizer	Overall	P valueb
No. of calls
PND 2	256 (317)	352 (472)	29 (36)	79 (94)	106 (143)	87 (164)	168 (291)	<0.001
PND 5	152 (157)	543 (499)	159 (150)	205 (506)	121 (133)	183 (155)	238 (352)	0.005
PND 8	390 (334)	567 (385)	301 (239)	177 (189)	402 (459)	239 (284)	349 (343)	0.013
PND 11	N/A	221 (282)	120 (132)	161 (149)	191 (171)	161 (144)	171 (181)	0.700
PND 14	N/A	235 (304)	202 (232)	140 (221)	241 (487)	211 (267)	206 (315)	0.800
Overall	263 (289)	401 (420)	176 (199)	157 (277)	221 (342)	182 (217)	234 (313)	<0.001
Younger	263 (289)	493 (452)	187 (207)	160 (330)	226 (331)	180 (223)	260 (339)	<0.001
Older	N/A	227 (286)	161 (190)	151 (183)	216 (360)	185 (210)	188 (255)	>0.900
Duration (s)
PND 2	0.014 (0.015)	0.011 (0.011)	0.008 (0.008)	0.013 (0.013)	0.012 (0.012)	0.010 (0.010)	0.012 (0.013)	<0.001
PND 5	0.009 (0.010)	0.010 (0.011)	0.008 (0.008)	0.012 (0.012)	0.009 (0.009)	0.008 (0.008)	0.010 (0.010)	<0.001
PND 8	0.012 (0.014)	0.009 (0.010)	0.009 (0.010)	0.010 (0.011)	0.010 (0.013)	0.012 (0.013)	0.010 (0.012)	<0.001
PND 11	N/A	0.007 (0.009)	0.008 (0.010)	0.009 (0.011)	0.010 (0.013)	0.008 (0.010)	0.009 (0.011)	<0.001
PND 14	N/A	0.006 (0.009)	0.006 (0.008)	0.007 (0.010)	0.009 (0.014)	0.007 (0.009)	0.007 (0.010)	<0.001
Overall	0.012 (0.014)	0.009 (0.010)	0.008 (0.009)	0.010 (0.011)	0.010 (0.012)	0.009 (0.011)	0.010 (0.011)	<0.001
Younger	0.012 (0.014)	0.010 (0.011)	0.008 (0.009)	0.011 (0.012)	0.010 (0.012)	0.010 (0.011)	0.010 (0.012)	<0.001
Older	N/A	0.007 (0.009)	0.007 (0.009)	0.008 (0.011)	0.010 (0.013)	0.008 (0.010)	0.008 (0.010)	<0.001
Mean USV (kHz/100)
PND 2	681 (123)	719 (128)	745 (109)	743 (104)	717 (104)	708 (154)	709 (127)	<0.001
PND 5	701 (185)	730 (138)	746 (161)	757 (175)	716 (141)	761 (164)	736 (157)	<0.001
PND 8	737 (168)	733 (149)	734 (152)	729 (141)	727 (155)	743 (154)	734 (154)	<0.001
PND 11	N/A	774 (146)	734 (151)	745 (155)	726 (146)	759 (151)	749 (150)	<0.001
PND 14	N/A	725 (143)	727 (129)	715 (152)	745 (142)	730 (129)	730 (139)	<0.001
Overall	712 (160)	733 (142)	735 (148)	740 (156)	729 (146)	744 (152)	732 (149)	<0.001
Younger	712 (160)	729 (141)	738 (153)	746 (156)	724 (147)	744 (158)	730 (150)	<0.001
Older	N/A	751 (147)	729 (137)	732 (154)	736 (144)	743 (140)	739 (145)	<0.001

Overall, all groups or time points combined; younger, PND 2, 5, and 8; older, PND 11 and 14.

^a All table cells depict mean (SD).

^b Kruskal-Wallis rank sum test.

Frequencies.

For older pups, there were no significant differences in USV mean frequencies between experimental groups at each time point presented in Figure 1A. Over time, all experimental groups’ T120 frequencies were significantly (P < 0.05) higher than their frequencies at other time points, except baseline (T0) (Figure 2A-E). Handling only baseline mean frequencies were significantly different compared with T120 mean frequencies (Figure 2A). For younger pups, hypothermia anesthesia yielded significantly lower (P < 0.001) USV mean frequencies compared with handling only during recovery and at 5 minutes postrecovery, as presented in Figure 1C. Similarly, hypothermia had significantly lower (P < 0.05) USV mean frequencies compared with isoflurane drop anesthesia at the same time points (Figure 1C). Over time, hypothermia and sevoflurane drop groups had significantly different (P < 0.05) USV mean frequencies between each time point (Figure 3B and F).

Figure 1.

Predicted mean USV frequencies and durations for pups at study points. (A) Predicted mean USV frequencies for older pups (PND 11 and 14). No significant differences were found between the experimental groups at any time point. (B) Predicted mean duration of USV calls for older pups (PND 11 and 14). Longer USV calls were observed in the sevoflurane drop group compared with the isoflurane vaporizer and handling only groups during and after recovery. (C) Predicted mean USV frequencies for younger pups (PND 2, 5, and 8). Decreases in mean USV frequencies were observed during recovery and after recovery in the hypothermia group when compared with handling only and isoflurane drop. (D) Predicted mean duration of USV calls for younger pups (PND 2, 5, and 8). Longer USV calls were observed in the hypothermia group compared with the isoflurane drop and sevoflurane vaporizer groups during and after recovery. Data are presented as mean ± SD. See Table 1 for the number of animals in each group. DR, during recovery; PR, 5 min postrecovery; T0, baseline; T120, 120 min postrecovery. *, P < 0.05; **, P < 0.01; ***, P < 0.001. (Continued)

Figure 2.

Predicted Mean USV Frequencies and Predicted Mean Duration for Older Pups (PND 11 and 14) by Group Over Time Points. USV frequencies were higher at T120 when compared with all other time points across all experimental groups: (A) handling only, (B) Isoflurane vaporizer, (C) isoflurane drop, (D) sevoflurane vaporizer, and (E) sevoflurane drop. A significant decrease in USV frequencies was observed during recovery and after recovery in the (C) isoflurane drop and (E) sevoflurane drop groups. Longer during and postrecovery USV calls were observed for (H) isoflurane drop and (J) sevoflurane drop group compared with their baseline calls, while no differences were observed in (F) handling only, (G) isoflurane vaporizer, and (I) sevoflurane vaporizer. Data are presented as mean ± SD. See Table 1 for the number of animals in each group. DR, during recovery; PR, 5 min postrecovery; T120, 120 min postrecovery;. T0, baseline *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Figure 3.

Predicted Mean USV Frequencies and Predicted Mean Duration for Younger Pups (PND 2, 5, and 8) by Group Over Time Points. Differences in mean USV frequencies were observed at each time point for (B) hypothermia and (E) sevoflurane vaporizer groups, while no differences were observed in (A) handling only, (C) isoflurane vaporizer, (D) isoflurane drop, and (F) sevoflurane drop. Longer during and after recovery USV calls were observed for (H) hypothermia compared with all other time points. No differences were observed in (G) handling only, (I) isoflurane vaporizer, (J) isoflurane drop, (K) sevoflurane vaporizer, and (L) sevoflurane drop. Data are presented as mean ± SD. See Table 1 for the number of animals in each group. DR, during recovery; PR, 5 min postrecovery; T0, baseline; T120, 120 min postrecovery. *, P < 0.05; **, P < 0.01; ***, P < 0.001; ****, P < 0.0001.

Duration.

For older pups, the sevoflurane drop group demonstrated significantly (P < 0.01) longer duration of calls compared with handling only during recovery and at 5 minutes postrecovery (Figure 1B). The sevoflurane drop group also had significantly (P < 0.05) longer duration of calls compared with isoflurane vaporizer, during recovery and at 5 minutes postrecovery, and sevoflurane vaporizer at 5 minutes postrecovery (Figure 1B). Over time, the duration of calls was significantly (P < 0.01) different within the sevoflurane drop group (Figure 2J). Significant differences (P < 0.05) in duration of calls were also seen in isoflurane drop over time, excluding T120 (Figure 2H). For younger pups, during recovery and at 5 minutes postrecovery, hypothermia anesthesia resulted in significantly (P < 0.05) longer duration of calls than handling only, isoflurane drop, and sevoflurane vaporizer (Figure 1D). In addition, isoflurane vaporizer anesthesia resulted in a significantly longer duration of calls compared with hypothermia during baseline. Of note, hypothermia had significant (P < 0.05) differences over time (Figure 3H).

Number of calls.

Within older pups, there were no significant differences in the number of calls between experimental groups or ages. Within younger pups, there were no significant differences between groups, but PND 8 had a significantly (P < 0.05) higher number of calls compared with PND 2 and 5.

Time to surgical plane of anesthesia and time to recover.

Within both age groups, isoflurane drop and sevoflurane drop time to surgical plane of anesthesia was significantly (P < 0.001) shorter than all other anesthesia groups but not significantly different from each other (Figure 4). In addition, within older pups, isoflurane vaporizer time to surgical plane of anesthesia was significantly different (P < 0.05) than sevoflurane vaporizer (Figure 4). For time to recover from anesthesia, there were no significant differences between all experimental groups in older pups (Figure 5). Within younger pups, time to recovery from anesthesia was significantly (P < 0.05) longer in hypothermia than all other anesthesia groups. Isoflurane vaporizer also had a significantly (P < 0.05) longer time to recovery from anesthesia compared with sevoflurane vaporizer and drop group in younger pups (Figure 5).

Figure 4.

Induction Time Measures Time to Surgical Plane of Anesthesia. Shorter induction times were observed in isoflurane drop and sevoflurane drop groups in both younger (PND 2, 5, and 8) and older (PND 11 and 14) pups. Data are presented as mean ± SEM. See Table 1 for the number of animals in each group. *, P < 0.001 compared with all other groups among older pups, except other drop method; **, P < 0.001 compared with all other groups among younger pups, except other drop method; +, P < 0.05 compared with isoflurane vaporizer among older pups.

Figure 5.

Recovery Time Measures Time for Righting Reflex to Return. Longer recovery times were observed in the hypothermia group in younger (PND 2, 5, and 8) pups. No significant differences in recovery times were observed in older pups. Data are presented as mean ± SEM. See Table 1 for the number of animals in each group. *, P < 0.05 compared with all other groups among younger pups; +, P < 0.05 compared with sevoflurane vaporizer and sevoflurane drop among younger pups.

Blood biomarkers.

Table 3 summarizes the mean glucose, corticosterone, and norepinephrine of neonatal mice in this study.

Table 3.

Descriptive Statistics for Biomarker Data

Outcomea	No handling	Handling only	Hypothermia	Isoflurane drop	Isoflurane vaporizer	Sevoflurane drop	Sevoflurane vaporizer
Glucose (mg/dL)
PND 2	110.8 (4.7)	109.8 (3.7)	112.8 (3.6)	111.2 (3.4)	113.8 (3.6)	102.2 (4.8)	104.3 (7.0)
PND 5	142.8 (3.7)	134.8 (2.6)	162.7 (9.7)	132.5 (3.5)	135.1 (4.1)	139.0 (2.4)	141.9 (5.9)
PND 8	161.8 (3.7)	152.9 (2.8)	170.2 (3.0)	156.0 (1.9)	150.5 (3.1)	145.1 (1.9)	151.8 (2.5)
PND 11	150.4 (4.5)	144.2 (5.0)	N/A	145.0 (2.5)	146.7 (6.2)	150.0 (4.1)	146.0 (3.8)
PND 14	165.9 (5.9)	151.3 (4.0)	N/A	137.7 (10.4)	154.7 (5.8)	147.7 (3.4)	150.2 (4.7)
Overall	145.9 (3.0)	138.5 (2.4)	148.0 (5.4)	136.9 (2.8)	140.0 (2.7)	137.6 (2.5)	138.9 (3.0)
Younger	138.7 (3.9)	133.8 (3.1)	148.1 (5.4)	134.2 (3.2)	133.2 (3.1)	129.4 (3.5)	133.7 (4.2)
Older	157.1 (3.8)	147.5 (3.3)	N/A	141.4 (5.3)	150.4 (4.3)	148.8 (2.6)	148.0 (3.0)
Corticosterone (ng/mL)
PND 2	25.7 (15.0)	26.8 (16.0)	30.9 (17.1)	16.2 (2.6)	49.6 (22.8)	57.3 (24.1)	34.8 (13.0)
PND 5	25.4 (17.3)	37.5 (20.1)	26.7 (3.6)	15.0 (3.5)	18.8 (6.5)	27.1 (17.2)	54.4 (22.4)
PND 8	25.8 (14.0)	62.0 (24.4)	21.0 (3.4)	31.3 (14.8)	51.8 (22.7)	30.3 (17.0)	18.4 (1.7)
PND 11	39.8 (13.5)	19 (2.7)	N/A	31.4 (5.8)	34.7 (4.7)	41.7 (15.2)	31.1 (4.1)
PND 14	57.1 (11.3)	53.0 (10.9)	N/A	70.3 (16.6)	60.5 (7.6)	80.9 (10.8)	62.6 (9.5)
Overall	34.1 (6.5)	40.2 (8.2)	26.3 (5.9)	32.6 (5.2)	42.4 (6.7)	47.7 (7.8)	38.7 (5.5)
Younger	25.7 (8.7)	43.2 (12.2)	26.3 (5.9)	39.6 (10.7)	39.6 (10.7)	37.8 (11.2)	34.4 (8.0)
Older	47.3 (9.1)	34.7 (6.1)	N/A	46.7 (4.9)	46.7 (4.9)	61.3 (9.9)	46.2 (5.9)
Norepinephrine (pg/mL)
PND 2	629.4 (94.4)	1,070.4 (266.3)	1,162.4 (248.8)	1,068.3 (288.9)	783.0 (135.5)	947.6 (316.3)	517.2 (61.6)
PND 5	795.1 (132.6)	753.3 (104.3)	732.7 (66.6)	888.3 (210.1)	748.9 (86)	576.4 (63.5)	819.3 (152.0)
PND 8	813.0 (94.1)	857.7 (143.6)	700.6 (98.9)	771.9 (136.9)	735.8 (91.2)	674.0 (93.9)	727.6 (93.5)
PND 11	1,536.1 (223)	1,324.1 (201.1)	N/A	1,089.8 (123.0)	1,164.6 (203.2)	1,226.3 (248.8)	1,326.8 (209.9)
PND 14	1,601.6 (294.8)	1,378.4 (170.7)	N/A	993.7 (54.9)	1,222.9 (234.2)	1,459.7 (197.3)	1,357.9 (304.2)
Overall	1,071.1 (91.8)	1,050.9 (84.6)	880.9 (103)	963.3 (80.9)	927.6 (73.6)	999.4 (99.6)	926.4 (84.1)
Younger	742.2 (61.6)	889.4 (103)	880.9 (103)	911.0 (127.4)	755.7 (59.7)	735.2 (112.3)	687.8 (62.3)
Older	1,564.5 (176.6)	1,349.2 (131.4)	N/A	1,041.8 (66.7)	1,191.6 (151.1)	1,343.0 (157.5)	1,341.8 (178.9)

Overall, all groups or time points combined; younger, PND 2, 5, and 8; older, PND 11 and 14.

^a All table cells depict mean (SEM).

Glucose.

Blood glucose levels in the no handling group were significantly (P < 0.05) higher than in mice anesthetized by the isoflurane drop and sevoflurane vaporizer methods in older pups (PND 11 and 14) (Figure 6). Hypothermia-anesthesia resulted in significantly (P < 0.05) higher blood glucose levels compated to sevoflurane drop anesthesia in younger pups (PND 2, 5, and 8) (Figure 6).

Figure 6.

Blood Glucose Levels for Younger Pups (PND 2, 5, and 8) and Older Pups (PND 11 and 14) Across Experimental Groups at Baseline (For No Handling Group) or 120 min After Recovery (For All Other Experimental Groups). In younger pups, hypothermia was the only anesthesia group with elevated glucose levels. In older pups, no significant differences were seen in glucose levels between anesthesia groups. Data are presented as mean ± SEM. See Table 1 for the number of animals used in each group. +, Value is significantly (P < 0.05) different from that of isoflurane drop among older pups (PND 11 and 14); *, value is significantly (P < 0.05) different than that of sevoflurane drop among younger pups (PND 2, 5, and 8).

Serum corticosterone.

Serum corticosterone levels in older pups that underwent handling only were significantly (P < 0.05) higher than those in isoflurane vaporizer and sevoflurane drop groups (Figure 7). Similarly, within younger pups, handling only resulted in significantly (P < 0.05) higher serum corticosterone levels compared to all other anesthesia groups, except sevoflurane drop. Younger pups that underwent no handling had significantly (P < 0.05) lower serum corticosterone levels than did both vaporizer anesthesia (isoflurane and sevoflurane) and hypothermia anesthetized pups. In addition, in younger pups, sevoflurane vaporizer serum corticosterone levels were significantly (P < 0.05) different than those for mice anesthetized by either of the drop methods (isoflurane and sevoflurane) (Figure 7).

Figure 7.

Serum Corticosterone Levels for Younger Pups (PND 2, 5, and 8) and Older Pups (PND 11 and 14) Across Experimental Groups at Baseline (For No Handling Group) or 120 min After Recovery (For All Other Experimental Groups). No significant differences were seen in corticosterone levels between anesthesia groups in older pups. No significant differences were seen in corticosterone levels between anesthesia groups in younger pups, except for sevoflurane vaporizer. Data are presented as mean ± SEM. See Table 1 for the number of animals used in each group. *, P < 0.05 compared with hypothermia, isoflurane vaporizer, and sevoflurane vaporizer among younger pups; **, P < 0.05 compared with hypothermia, isoflurane vaporizer, sevoflurane vaporizer, and isoflurane drop; +, P < 0.05 compared with isoflurane vaporizer and sevoflurane drop among older pups; ^, P < 0.05 compared with no handling, handling only, isoflurane drop, and sevoflurane drop among younger pups.

Serum norepinephrine.

In older pups, no handling and handling only resulted in serum norepinephrine levels that were significantly (P < 0.05) different than levels in mice that were anesthetized by isoflurane vaporizer, isoflurane drop, and sevoflurane vaporizer, as shown in Figure 8. Older pups anesthetized by sevoflurane drop had significantly (P < 0.05) different serum norepinephrine levels compared with all other anesthesia groups. Lastly, isoflurane drop anesthesia resulted in significantly (P < 0.05) higher serum norepinephrine levels than did all other anesthesia groups in younger pups (Figure 8).

Figure 8.

Serum Norepinephrine Levels for Younger Pups (PND 2, 5, and 8) and Older Pups (PND 11 and 14) Across Experimental Groups at Baseline (For No Handling Group) or 120 min After Recovery (For All Other Experimental Groups). In younger pups, all experimental groups except isoflurane drop showed similar norepinephrine levels. In older pups, sevoflurane drop was the only anesthesia group with elevated norepinephrine levels. Data are presented as mean ± SEM. See Table 1 for the number of animals used in each group. +, P < 0.001 compared with isoflurane vaporizer, sevoflurane vaporizer, and isoflurane drop among older pups; *, P < 0.05 compared with hypothermia, isoflurane vaporizer, sevoflurane vaporizer, and sevoflurane drop among younger pups; **, P < 0.05 compared with isoflurane vaporizer, sevoflurane vaporizer, and isoflurane drop among older pups.

Discussion

Using ultrasonic vocalization as an indicator of distress and physiologic stress in neonatal mice, we evaluated the acute effects of different methods of anesthesia. We found that pups anesthetized with hypothermia experienced heightened levels of acute distress, evidenced by their significantly lower frequencies and longer duration of calls during, and immediately after, recovery from the anesthetic event. Hypothermic pups also had longer recovery times, leading to prolonged dam separation and stress. These findings support our hypothesis and suggest that hypothermia is the least preferred anesthetic method for neonatal mice when compared with inhalant anesthesia methods.

Historically, hypothermia has been used as a method of anesthesia for neonatal rodents <10 days of age,¹⁰ due to their neurologically immature and poikilothermic status. Studies have presented that brief hypothermia anesthesia of neonatal mice causes no long-term effects on body weight, serum corticosterone levels, or adult behavior.^14,34 To our knowledge, this is the first study to investigate the acute effects of hypothermia anesthesia on neonatal mice. Consistent with prior findings in neonatal rats,³¹ we observed longer duration of ultrasonic vocalization, as well as lower frequencies associated with neonatal mice that underwent hypothermia anesthesia. These vocalization characteristics have been associated with a negative affective state and alterations of breathing patterns in rodents.³⁵

The heightened acute distress observed in hypothermia-anesthetized pups during and immediately after recovery may stem from sudden pain or tissue injury from the extreme drop in body temperature. No acute histologic skin changes have been reported in neonates euthanized by hypothermia.³⁴ Thus, the source of pain or skin injury in neonatal mice could be attributed to the rewarming process after hypothermia anesthesia. Although incubators could help reduce rewarming injuries,¹¹ these devices are not as readily accessible as other heating alternatives such as heating pads. Therefore, the exposure to a potential pain source (rewarming) is exacerbated as a result of prolonged recovery time, reportedly ∼20 minutes in neonatal hypothermia rodents.^15,31

A suitable alternative to hypothermia neonatal anesthesia is inhalant anesthesia. Our results show that inhalant anesthesia was less stressful than hypothermia in mice <8 days old. No significant differences in USV inhalation data were observed between the inhalant anesthesia groups and the handling only group at PND 2-8. These findings suggest that inhalant anesthesia does not induce additional stress beyond the inherent stress associated with maternal separation during neonatal experimental procedures. Although all study deaths occurred in pups receiving inhalant anesthesia, the deaths spanned different age groups and represented a 4.5% mortality rate (9 of 197 pups anesthetized), which is comparable to another published study in neonatal rats undergoing inhalant anesthesia.³¹ Both inhalant anesthesia and hypothermia are considered to be acceptable with conditions as methods for euthanasia of neonatal rodents,¹⁰ which inherently introduces the risk of mortality when these techniques are used for anesthesia. This risk can be mitigated through shorter anesthesia durations, user experience, and reduced recovery times. Despite similar induction times, inhalation anesthesia showed faster recovery times than hypothermia anesthesia.

There was no consistent increase in blood biomarkers within any anesthetic group in the younger pups. Norepinephrine, corticosterone, and glucose are released as part of the stress response activated by the nervous system and the hypothalamic–pituitary–adrenal axis.³⁶ Norepinephrine is released first, within 5-30 seconds of a stressor, followed by corticosterone and glucose within 4 minutes.³⁷ The duration of elevated biomarker levels varies depending on the severity of the stressor. Importantly, these biomarkers can also be influenced by nonstress-related factors. For example, blood collection after a cold stress (hypothermia) via decapitation has been reported to result in lower norepinephrine levels comparable to those obtained via venipuncture.³⁸ In this study, any resulting biomarker changes may have been masked by the blood collection method (decapitation). Alternative blood collection methods are seldom used in neonatal rodents due to their very limited blood volume.

In addition, neonatal age and maternal care are known to suppress stress responses and reduce the release of corticosterone and norepinephrine in neonatal rodents.³⁶ Throughout this study, dams were consistently observed nesting and nursing their pups despite frequent pup handling, potentially reducing sustained elevations in blood biomarkers. Although glucometers have limited validity and are generally recommended for clinical management rather than research application,³⁹ their use in neonates is advantageous because only a small blood volume is required. Given the multifactorial influences on blood variability and the immaturity of neonatal stress pathways, developing novel methods to assess stress in neonatal rodents is essential. Behavioral assessments, specifically USVs, may prove beneficial as increasing studies are standardizing technological protocols and rodent vocalization ranges.^{23,27,28,31,40–42} Based on our findings, we strongly support replacing hypothermia with inhalant anesthesia to improve animal welfare, in accordance with the 3Rs principles.⁴³

Finally, older neonatal pups (PND 11 and 14) exposed to the sevoflurane drop method experienced heightened levels of distress, as evidenced by significantly longer duration of calls, elevated corticosterone, and elevated norepinephrine blood levels. Of the 5 older pup deaths in this study, all of them underwent sevoflurane inhalant anesthesia. Older neonatal mice pups that undergo sevoflurane anesthesia may be more vulnerable to stress-related responses such as severe hypoglycemia when compared with other anesthesia inhalant methods,⁹ which may have contributed to the deaths in this group. Other potential causes of death include anesthetic overdose and natural inter-animal variability. These findings highlight the need for further investigation, given that sevoflurane is commonly used for neonatal mice anesthesia.

This study has several limitations. First, USV recordings could not be paused to allow neonatal pups stimulation before the 5-minute mark, which, in turn, may have prolonged recovery times. Second, injectable anesthetic protocols were not evaluated because of the high costs and limited availability of microdosing equipment, which also reduces the feasibility of their widespread use by researchers. Future studies should examine pain- or stress-related outcomes associated with injectable anesthesia in neonates. Third, the statistical model for the number of USV calls deviated from the assumption of normality and had below zero predictions, although these data are unlikely to alter the overall conclusions. Fourth, only one inbred mouse strain (C57BL/6) was used. Including additional commonly used strains in future research would improve generalizability and account for strain-specific differences. Heart rate, temperature, respiratory rate, and vapor pressures were not recorded due to limited access to specialized equipment, which we acknowledge as an additional study limitation. Finally, we did not perform analysis of the USV call structures described in other studies. Future investigations should incorporate advanced call categorization to better characterize neonatal communication and refine pain- and stress-related USV indicators.

In conclusion, USVs offer a sensitive and reliable indicator of neonatal stress, particularly in pups <8 days old, when traditional biomarkers, such as corticosterone and norepinephrine, are less responsive during the stress hyporesponsive neonatal period.⁴⁴ These findings underscore the importance of refining neonatal mice anesthesia and providing better animal welfare, by selecting inhalation neonatal anesthesia, especially as specialized equipment, standardized protocols, and real-time physiologic monitoring become more accessible and reliable.

Conflict of Interest

The authors have no conflicts of interest to declare.

Funding

This work was supported by the ACLAM Foundation grant and investigator start-up funds (to M.D.W.). The funders of the study had no role in the study design, data collection, data analysis, data interpretation, or writing of the manuscript.

Generalizability/translation

The findings of this study on hypothermia have now been presented to produce similar results in neonatal mice and rats. These results can likely be generalized to other neonatal altricial rodents; however, additional species-specific studies are needed for confirmation. The use of USV is common across various species and experimental conditions.

Protocol registration

The experimental protocol was prepared prior to the start of study and registered at the University of North Carolina-Chapel Hill.

Data availability

Data supporting the conclusions of this study will be made available by the authors upon request.