Techniques used to directly measure pulmonary mechanics in mice represent the "gold standard", but generally require anesthesia, intubation and expertise in handling.

The classical approach to determine lung function in mice is the measurement of pulmonary resistance (R_L) and dynamic compliance (C_dyn) in response to non-specific bronchoconstrictors. In 1988 Martin et al. demonstrated the feasibility of R_L and C_dyn measurement in anesthetized, tracheotomized and mechanically ventilated mice 14. To assess R_L and C_dyn determination of transpulmonary pressure and flow are required. In mice the chest wall has been shown to present little mechanical load compared to the mechanical load of the lung 15, unless there is some pathology of the chest wall. Thus direct measurement of transpulmonary pressure is generally not mandatory 16. Tidal flow is commonly derived from the differentiation of the volume signal. R_L and C_dyn can then be calculated by fitting an equation of motion to the measurements of pressure, flow and volume 4. In this equation, P_TP = V × R_L+ VT/C_dyn, P_TP is transpulmonary pressure (or in the mouse ≈ transrespiratory pressure), V is tidal airflow, R_L is pulmonary resistance, V_T is tidal volume, and C_dyn is the dynamic pulmonary compliance. The invasive measurement of R_L and C_dyn by body plethysmography normally requires surgical instrumentation of the trachea in anesthetized animals. It is common to use pentobarbital sodium (70–90 mg/kg) administered intraperitoneally as anesthetic because it normally provides an adequate depth of anesthesia for at least 30 minutes. Alternative anesthetic regimens in mice have been described 67. It is important not to disturb and agitate the animal beforehand, as this may impact the quality of the subsequent measurement. Useful reflexes to ensure that an adequate depth of anesthesia has been attained include loss of the righting reflex (lost during the onset of anesthesia) and of the toe-pinch reflex (lost during medium to deep anesthesia). If the animal attempts to withdraw its limb, then it is not sufficiently anesthetized and should be administered an additional dose (~10–20% of the initial dose).

Determination of R_L and C_dyn not only provides the classical determination of airway responsiveness, but also provides a more detailed insight into pulmonary mechanics. R_L reflects both narrowing of the conducting airways and parenchymal viscosity. In contrast C_dyn is considered to primarily reflect the elasticity of the lung parenchyma, but is also influenced by surface tension, smooth muscle contraction and peripheral airway inhomogeneity. Numerous methods for determining R_L and C_dyn have been described in anesthetized and instrumented mice 457. One option is to use a (mass-constant) body plethysmography box with the tracheal cannula leading out of the plethysmograph 1718. When mechanical ventilation is indicated, tracheostomy is usually performed for endotracheal intubation of the deeply anesthetized animal. The surgically exposed trachea is viewed directly and the incision is made in the upper third of the trachea to allow proper insertion of the cannula and to avoid measuring artifacts.

The tracheostomy tube can then be attached to a four-way connector, where two ports of the connector are attached to the inspiratory and expiratory sides of a ventilator and the remaining tube to a pressure transducer that measures tracheal pressure. Ventilation should then be set at a rate comparable to normal breathing (around 150 breaths/min, tidal volume ≈ 8–10 ml/kg) with a positive end-expiratory pressure (PEEP) of 2–5 cm H₂O. It is important to use PEEP in mice even with the chest closed, since functional residual capacity (FRC) in conscious mice is normally maintained with active inspiratory muscle tone that is minimal or eliminated in the anesthetized animal 16. Lung volume changes must be assessed by calibrating the plethysmographic pressure. To stabilize the volume signal for thermal drift the body plethysmograph chamber can be connected to a large bottle filled with copper gauze.

To assess airway responsiveness, cholinergic bronchoconstrictive agents such as methacholine (MCh) are administered to the animal at increasing doses either by aerosol inhalation or systemically by intravenous administration via the tail or jugular vein. Airway responsiveness is assessed either as the change in R_L compared to baseline or as the peak response after challenge. Before each series of challenge doses the lung should be briefly hyperinflated to standardize the volume history. Measurements are made of the absolute values of the responses of C_dyn and R_L and as a percentage of baseline, determined from an initial vehicle challenge.

The key advantage of the invasive approach is the reproducible and precise assessment of transient changes in pulmonary mechanics in mice. The insertion of a tracheal tube also avoids measurement of changes in the upper airways, and provides the opportunity for taking broncho-alveolar lavage (BAL) samples after lung function measurements. Disadvantages of conventional invasive measurements include surgical tracheostomy thus precluding repeated measurements, the need for anesthesia, mechanical ventilation and expertise in handling.

As outlined above, the utility of invasive determination of murine lung function is generally limited by several factors. Recent methodological advances, however, have improved the ability to measure lung mechanics on repeated occasions 19. These modifications involving direct laryngoscopy have now enabled repetitive determination of pulmonary mechanics (R_L and C_dyn) in combination with local aerosol administration via an orotracheal tube in intact animals 91020.

With one of these approaches, intubation is done with a standard 20G × 32 mm (1 1/4 inch) teflon cannula (e.g. Abbocath^®-T cannula, Abott, Ireland) in anesthetized mice that are suspended by their upper incisors from a rubber band and the midthorax held by an elastic band on a 65° incline Plexiglas support to facilitate intubation. We have made positive experience using anesthesia plus analgesia with 20–30 mg/kg etomidate and 0.05 mg/kg fentanyl given intraperitoneally (i.p) with minimal supplementations as required or volatile anesthesia with halothane 1.5 % plus propofol 70 mg/kg i.p. Paralysis is not mandatory. A metal laryngoscope (length 12 cm plus an additional 1.8 cm at an angle of 135°, with 0.3 cm) is used as a tool to allow visualization of the tracheal opening which is transilluminated below the vocal cords by a halogen light source. The direct visualization of the trachea allows gentle insertion of the cannula into the tracheal opening 1921. Orotracheal intubation of the anesthetized mouse takes about five minutes and has also been successfully applied in mouse cardiac surgery 21. Alternatively, a Seldinger technique has been described using a 0.5 mm optical light fiber as an introducer over which the cannula is slid down into the proximal trachea 22. The intubated, spontaneously breathing animal is then placed in supine position in a thermostat-controlled whole-body plethysmograph (Figure 1). The orotracheal tube is directly attached to a pneumotachograph/differential pressure transducer unit to record tidal flow. To measure transpulmonary pressure (PTP), a water-filled polyethylene (PE)-90 tubing is inserted into the esophagus to the level of the midthorax and attached to a pressure transducer. R_L and C_dyn are calculated over a complete respiratory cycle with an integration method over flows, volumes and pressure 1023. The resistance of the orotracheal tube (0.63 cm H₂O·s·ml^-1) is subtracted from R_L recordings.

This approach was validated in several groups of BALB/c mice 10. The results showed that dose-related increases in R_L and C_dyn to inhaled cholinergic challenge with MCh were reproducible over short and extended intervals without causing significant cytological alterations in the BAL fluid or relevant histological changes in the proximal trachea and larynx regardless of the number of orotracheal intubations.

A key advantage of this method which combines orotracheal intubation via direct laryngoscopy and local administration of aerosols directly into the lung is the repetitive assessment of classical measures of pulmonary mechanics to defined inhalation challenges in intact individual mice. Because the orotracheal cannula is tapered, a tight seal develops as it is inserted into the proximal trachea. This enables use of this method in spontaneously breathing as well as in mechanically ventilated mice. Orotracheal intubation further offers the opportunity to collect BAL samples in vivo on multiple occasions in the same animals 24. Limitations include the need for anesthesia, instrumentation of the trachea and expertise in handling.

Another approach for invasive assessment of airway function in mice is the low-frequency forced oscillation technique (LFOT). The LFOT was derived from similar techniques used in humans and larger animals and produces estimates of lung impedance (Z) which can be considered the most detailed measurement of pulmonary mechanics currently available 48112526.

Different parts of the impedance frequency spectrum reflect different parts of the respiratory system. Impedance data can be further analyzed using the Constant Phase Model which provides a suitable assessment of pulmonary mechanics 27. Fitting the Constant Phase Model to oscillatory data allows airway and tissue mechanical components to be distinguished.

Until recently, little was known about lung impedance of mice, particularly because of technical difficulties of measuring lung impedance precisely. Lung impedance consists of two parts. One part of impedance, resistance (R), describes essentially the resistance of the conducting airways (Raw) and tissue (Rti). The second part of impedance, referred to as reactance (X), reflects respiratory compliance (1/elastance) and characterizes the lung parenchyma. The contribution of the inertance (I) of the gas in the murine airways, however, is only significant at frequencies ≥ 20 Hz. The main advantage of this approach for measuring lung function, compared to the classical methods of assessing airway resistance and dynamic compliance, is that the more sophisticated mathematical models may better represent the complexity of the intact lung.

Two different methods have been developed to assess lung impedance in small animals. One technique uses a small plastic wave tube that is placed into the trachea and is attached to a loudspeaker 428.

The properly miniaturized wave tube has a precisely known geometry and material constant. During the measurement ventilation is paused and the setting is switched from the ventilation to the measurement circuit. The loudspeaker produces an oscillatory flow through the tube and lung impedance is assessed from flow and pressure measurements along the tube. From the pressure spectra along the tube lung impedance can be assessed 29. This technique is particularly useful for the precise measurement in very young mice, where other techniques such as the piston pump oscillator may be critical 30. The second method uses a computer-controlled piston pump. This system not only allows for mechanical ventilation of the animal but also for precise frequency and amplitude control of the applied oscillations. The constant phase model is then fit to the data obtained from the multiple frequencies simultaneously applied at the airway opening, thereby enabling determination of the airway and lung tissue impedance. This model involves three independent variables: airway resistance (R) as a marker of central airway resistance, tissue damping (G) is related to tissue resistance and reflects the dissipative properties, while tissue elastance (H) describes the elastic properties of the lung tissue.

LFOT correlates well with classical measures of lung resistance and has been successfully used to assess airway responsiveness in mouse models of allergic airway disease 4152528313233. The computer-controlled ventilator also allows the assessment of quasi-static compliance. As with other invasive techniques, the animals need to be anesthetized, tracheally intubated and then connected to the computer-controlled ventilator (e.g. set at a rate of 150 breaths per minute and a tidal volume of 10 ml/kg), with application of 2–5 cm H₂O PEEP. Mice can then be challenged with bronchoconstrictors by inhalation or via intravenous routes. It should be considered that while LFOT can be employed during apnea only, paralysis is not mandatory in anesthetized mice.

The main advantage of this technique is the detailed analysis of airway function and particularly the clear distinctions between central airways and more peripheral changes. This approach, however, also shares similar disadvantages with other invasive techniques as shown in Table 1. In addition, at least for assessing airway hyperresponsiveness it is still unclear what additional value lung impedance recordings provide over simpler measures of pulmonary mechanics 34.

Noninvasive plethysmographic methods of monitoring pulmonary function are preferred for long-term serial study designs as well as for screening large numbers of conscious mice. In many instances, a combination of invasive and noninvasive techniques is required to fully understand the physiologic significance of a respiratory phenotype.

In barometric whole-body plethysmography mice are placed in a closed chamber and the pressure fluctuations that occur during the breathing cycle are recorded 35. In contrast to invasive measurements of airway function animals are neither anesthetized nor instrumented and are relatively unrestrained. The major benefit of this noninvasive technique is that repetitive measurements can be done in the same mouse. Using a pressure transducer the pressure differences between the main chamber of the plethysmograph where the animal is placed and a reference chamber are assessed (Figure 2).

From this pressure time curve several parameters can be determined including breathing frequency, inspiratory and expiratory time as well as the maximum box pressure during inspiration and expiration. None of these variables is specific nor sensitive enough for being a suitable marker of airway responsiveness. From the box pressure signal during inspiration and expiration, and the timing comparison of early and late expiration, a dimensionless parameter called "enhanced pause" (Penh) has been calculated. It is notable that we do not refer to the as yet non-validated method of measuring Penh in freely moving mice.

To monitor responsiveness mice are exposed to a nebulized bronchoconstrictor such as MCh and changes in Penh are recorded for ~2–5 minutes for each aerosol challenge. Usually the response is expressed as fold increase of Penh for each MCh concentration compared with Penh values after an initial buffer challenge with the aerosolized vehicle.

Early studies in mice and other species showed a correlation between changes in Penh following methacholine challenge and lung function parameters determined by invasive lung function measurements and the technique has been widely used 1836373839. Based on this early work and because of the convenient handling of the animals, this method gained popularity in many research labs. An increasing amount of observations, however, have now cast doubt on the validity of Penh to reflect airway narrowing. Several reports found discrepancies in the amount of airway responsiveness when comparing Penh to conventional parameters of pulmonary mechanics 404142. Further evaluation of Penh demonstrated that events completely unrelated to lung mechanics such as humidification and warming of inspired gas, hyperoxia, and the timing of ventilation, have a major effect on the measurement 3141. These more careful and theoretical findings have thus led to a justifiable scepticism for using Penh as a reliable marker of airway obstruction 434445.

Nevertheless, in principle and consistent with current cautionary warnings, Penh may be useful for gross screening of overall lung function in small animals 43. Seen by itself, however, Penh says nothing about airway responsiveness and researchers who use it should corroborate the measurements with parallel, independent direct measurements of pulmonary mechanics 574445. Pros and cons of this method are summarized in Table 2.

Recent emphasis on the benefits of noninvasive technology has renewed interest in analyzing expiratory tidal flow patterns as a tool in the assessment of airway obstruction. Although noninvasive measurement of murine respiratory function has virtually become synonymous with the widely used barometric whole-body plethysmography method 35, some other noninvasive methods have been described 13464748.

The noninvasive measurement of midexpiratory flow (EF₅₀) as measured by head-out body plethysmography (Figure 3) was first described as an appropriate instrument to measure airway responsiveness in conscious mice by Alarie et al 48. With this method, airway constriction induces characteristic changes in the tidal flow pattern, which are best revealed by a decrease in tidal midexpiratory flow (EF₅₀, [ml/s]) (Figure 4). The change in EF₅₀ is typically linked with a reduction in tidal volume (VT), breathing rate (f) and prolonged expiratory time (TE). EF₅₀ can be determined with a glass head-out body plethysmography system. Animals are gently placed in the body plethysmographs while the head of each animal protrudes through a neck collar into a ventilated head exposure chamber. Aerosols can be delivered directly through the head exposure chamber. Tidal flow measurement is made with a calibrated pneumotachograph and a differential pressure transducer attached to the top port of each body chamber. The amplified and digitized flow signals are integrated with time to obtain tidal volume. From these signals several standard respiratory parameters, including tidal volume, breathing frequency, time of inspiration and expiration, and EF₅₀ can be derived from software analysis.

Validation studies in mice have demonstrated that the decline in EF₅₀ to inhaled cholinergic and allergic challenge closely reflects the decreases in simultaneously recorded pulmonary conductance (G_L = 1/R_L) and dynamic compliance (C_dyn) 10. The EF₅₀ method has been applied in several experimental situations, including animal models of experimental asthma, post-pneumonectomy, hyperoxia, and to study the effects of airborne toxicologic agents 313948495051525354. Advantages of this approach are its noninvasiveness and its allowing simple, rapid and repeatable measurements of several conscious animals at a time. Moreover, EF₅₀ is based on physiological principles and has a physical meaning [ml/s] that is directly related to airway resistance, thus enabling quantitative interpretation of airway changes between animals 55. In principle, head-out body plethysmography as described by Alarie et al. also enables evaluation of the sensory irritation potential of inhaled agents by recording the prolongation of the postinspiratory pause in mice 4851.

Concerns include the uncertainty about the potential contribution of upper airway resistance. To minimize effects of restraint stress on responses, monitoring of respiratory function should not be started until animals and individual measurements have settled down to a stable level. Because it has been shown that EF₅₀ may underestimate the magnitude of bronchoconstriction 910 it is still unclear how much this limits its use in detecting less pronounced changes in airway hyperresponsiveness. Accordingly, when such circumstances are present, EF₅₀ measurements should be confirmed by more direct assessments of pulmonary resistance. Table 3 summarizes the pros and cons of EF₅₀ measurements.

The author(s) declare that they have no competing interests.

TG and CT conceived of the review and drafted the manuscript, AB helped to draft the manuscript, WM helped to draft, discuss and revise the manuscript. All authors read and approved the final manuscript.