INTRODUCTION

Top Abstract Introduction References

THE INITIAL PREPARATION of isolated smooth muscle cells from the stomach of the toad Bufo marinus in 1971 (2, 13) made possible a number of major advances in our understanding of vertebrate smooth muscle. Studying isolated cells allows muscle responses to be measured without the influence of substances released from endothelial cells, epithelial cells (27), and nerve endings. Fay and collaborators used this preparation to make numerous important contributions in several areas, such as ultrastructure (22), contractile mechanisms (11, 37), Ca²⁺ movements (3, 40), membrane transport (24), and biochemical mechanisms regulating contraction (18). Later, several investigators isolated cells from mammalian smooth muscle tissues including ferret aorta (8), swine (10) and bovine (33, 39) carotid and coronary artery, rabbit intestine (4), and guinea pig stomach (6).

Isolated airway smooth muscle (ASM) cells have been prepared from several species (dog, rat, guinea pig, human), and the contractile responses of the canine cells have been studied (19). Whereas some investigators have succeeded in measuring force development by the single cells isolated from other smooth muscles (7, 38), force measurement of single cells is extremely difficult and requires expensive specialized equipment. In contrast, measuring the unloaded shortening velocity of a single smooth muscle cell is relatively easy and does not require the elaborate instrumentation required for single-cell force measurements.

In this study, we first describe a method for isolating viable, elongated, and Ca²⁺-tolerant muscle cells from both adult and preterm sheep airways. Then we describe an apparatus and method that allow unloaded shortening of the cells to be recorded in a perfusion chamber where the agonist is delivered by the constant flow of perfusion solution. Finally, we present a method for analyzing the shortening response of the isolated, unloaded cell that allows shortening velocities to be calculated from measurements of the cell length at various times on playback of a videotape.

Over the last several decades, considerable progress has been made in characterizing developmental (preterm, newborn, and adult) differences in tracheal physiology and mechanics (pressure-volume and pressure-flow relationships and the effects of mechanical ventilation; Refs. 5, 14, 17, 26, 28-31). These studies, with the use of the sheep and other species, lend valuable insight into how the very premature infant may differ from the later-term and full-term neonate with respect to airway function. We wish to build on these studies by learning about the mechanical properties of the ASM at the single-cell level.

	METHODS

Tissue

Solutions

MOPS-physiological salt solution (PSS). The composition of MOPS-PSS was (in mM) 117.8 NaCl, 0.027 Na₂EDTA, 6.0 KCl, 1.2 MgSO₄, 24.3 MOPS (pH 7.5), and 5.6 glucose. It was filtered through a 0.2-µm filter. This solution was nominally Ca²⁺ free.

Collecting solution. The collecting solution was identical to MOPS-PSS except it lacked glucose and contained both calcium (1.6 mM CaCl₂) and phosphate (1.2 mM NaH₂PO₄).

Perfusion solution for the flow chamber studies. The standard solution for flow chamber studies was MOPS-PSS with CaCl₂ added to a final concentration of 1.6 mM.

Agonist perfusion solution for the flow chamber studies. This solution was MOPS-PSS to which was added 100 µM agonist, 1% sucrose, 1 mg/ml Trypan blue, and 1.6 mM CaCl₂. These solutions were filtered (0.2 µm) before use.

Enzymes and Reagents

Dissection and Tissue Digestion Conditions

Papain (20 U/mg) was used at a concentration of 2 mg protein/ml in MOPS-PSS. DNase II (1 mg/ml) was also added. DTE (2 mM) was added to the digestion solution to activate the papain by reducing the sulfhydryl groups. The DTE must be fresh, or the papain will not be active or may not dissolve. DTE was added from a 0.2 M stock solution kept under a 100% N₂ atmosphere and sparged with N₂ after each use. The enzyme solution was warmed in a 37°C bath for 4-15 min until the papain dissolved and was then immediately used. Our laboratory (9) found previously that bubbling papain solutions with oxygen during digestion is unadvisable, probably because it interferes with the reduction of the sulfhydryl groups of papain.

The clean muscle was cut from the cartilage, added to the digestion solution, and placed in the 37°C bath for 20-60 min, depending on the amount of tissue, with an average of 30 min. Once the tissue appeared swollen and frayed, it was removed from the bath, rinsed gently with MOPS-PSS, and placed in a 90-mm-diameter plastic petri dish containing 20 ml MOPS-PSS and 2 mg DNase I. DNase I was added to the petri dish for better separation, settling, and collection of cells. Digestion with other enzymes (collagenase, dispase, elastase, and chymopapain) did not produce satisfactory results.

Individual Cells and Microscopy

Optical System

Recording Cell Contractions

A number of agonists were used, including caffeine, KCl, and cholinergic agonists (acetylcholine, bethanachol, carbachol, and methacholine). Carbachol was used in most experiments because it is more resistant to choline esterases.

Calculating Shortening Velocity

Cell lengths were fit to the following exponential model: length = L_min + (L_max  L_min) exp[v (time  lag)], where L_max is the longest length measured, L_min is the shortest length reached, time is the time elapsed since the switching valve started delivering agonist solution, lag is the time at which rapid contraction starts, and v is the exponential rate constant. These parameters are explained below.

The exponential rate constant v, with units of reciprocal seconds (s¹), is used as the shortening velocity. It represents velocity in cell lengths per second and allows shortening velocities of different sized cells to be compared. Lag is the earliest time point that is fit by the curve. The fitting program omits a few points at the beginning of the shortening phase, if necessary, to get the best fit. Because times are measured by the video stopwatch from the moment the switching valve is activated, "lag" has no physiological significance because it mostly represents the time required for agonist solution to flow from the switching valve into the dish and reach the cell and, therefore, is quite variable depending on flow rates and cell location. It is not critical to determine the exact instant when a cell starts to contract.

Usually the arrival of the agonist solution containing the Trypan blue dye was plainly visible, and it would have been possible to start the timer at the instant that the blue solution reached the cell. This would have given smaller values of lag that would represent only the time required for a cell to respond to a stimulus. Sometimes, though, the boundary between the clear and blue solutions was not sharp enough. Therefore, we always started the timer the moment the switching valve was activated, and the recorded videotape had the timer information on each video frame.

Accuracy of Cell Length and Velocity Measurements

The accuracy of the cell-length measurements depends on a number of factors, including the microscope and camera optics, the resolution of the digital image, the accuracy and resolution of the digitizing tablet used to trace the image of the cell, and the ability of a human operator to trace the image. The 512 × 400 pixel image allowed 0.25% resolution, and the digitizing tablet used to trace the cell images was capable of 0.15% resolution. After analyzing these factors, we concluded that the precision of the human operator tracing the image is likely to be the limiting factor in most measurements. Repeated measurements of the same cell are usually within 2% of each other. Measurement errors occurring if one end of the cell is out of the plane of focus are discussed in the DISCUSSION, but, taking these errors into account, we estimate that cell-length measurements are accurate to within 5% of the true cell length. It is important to use high optical magnification to maintain this accuracy, and we used a ×40 objective for the data reported here.

	RESULTS

Characteristics of the Cells

View larger version (41K): [in this window] [in a new window]   Fig. 1.   Tracheal smooth muscle cells isolated from adult and preterm sheep. Both images were obtained with ×40 differential interference contrast (Nomarski) optics by using an Olympus IMT-2 inverted microscope. Gray-scale images (8 bits) were captured directly to a computer file with a Dage-MTI Newvicon video camera. Top: smooth muscle cell from a preterm trachea; cell length is 121 µm. Bottom: smooth muscle cell from an adult trachea; cell length is 249 µm.

Nuclei are not visible in Nomarski, bright-field, or phase-contrast images of healthy adult cells. Nuclei in damaged or lysed adult and preterm cells are readily seen with bright-field or phase-contrast optics. However, in healthy, Ca²⁺-tolerant, preterm cells that contract in response to agonists, nuclei can sometimes be seen. Presumably, the smaller amount of contractile material in the preterm cells makes the nuclei easier to locate.

Average cell lengths, in 1.6 mM CaCl₂, for a sample of adult and preterm cells are given in Table 1. Adult cells were much longer than neonatal cells [194 ± 57 (SD) µm for adults and 93 ± 32 µm for neonates]. Preterm cells were also narrower, averaging 5 µm, compared with 8 µm wide for adult cells. These differences were statistically significant (P < 0.05). Despite their shorter length, preterm cells are long enough to contract substantially. However, it was more difficult to isolate cells from neonatal trachea, probably because the trauma is greater in the smaller, more fragile tissue. The cell volumes estimated from these mean length and width measurements are also shown, assuming two simple geometries, a cylinder and two cones joined at the base. Neither of these geometries is an exact description of the shape of the cell, but the true volume would be between these two estimates. The mean volume of the preterm cell is only about one-fifth the mean volume of the adult cell.

Contraction of Cells

Figure 2 depicts the contraction of an adult ASM cell. Figure 2A shows a plot of cell length (measured from still frames of videotaped contractions) as a function of time. Time was measured from the instant when the switching valve was activated, not the instant when the agonist reached the cell. Trypan blue in the agonist solution allowed the arrival of the agonist at the cell to be seen. Usually the cells contracted within 2-3 s of the arrival of agonist. In the example shown, the carbachol-Trypan blue solution reached the cell at 31 s, the cell was contracting at 33 s, and the contraction was complete by 41 s. Figure 2B shows images of this cell as it contracted from 322 to 85 µm.

View larger version (12K): [in this window] [in a new window]   View larger version (141K): [in this window] [in a new window]   Fig. 2.   Contraction of an adult ASM cell. A: plot of cell length as a function of time during shortening. Agonist solution was 100 µM carbachol, containing 1% sucrose to increase density (see METHODS) and 1 mg/ml Trypan blue so flow could be monitored. Flow rate was ~0.5 ml/min, and apparent linear velocity of agonist past cell was 0.6 mm/s. Agonist reached cell at a time corresponding to 31 s, denoted by downward arrow. Circles, measured length of smooth muscle cell at times indicated; curved line, best-fit line to these points (see Calculating Shortening Velocity). B: photomicrographs of same cell at selected points in contraction. Images were acquired with a ×40 objective and digitized. Composite image was created with the program NIH Image. Brightness and contrast of images were adjusted to compensate for dye mixed with agonist solution. First panel: cell at a time (23 s) before carbachol has reached it, cell length is 322 µm; second panel: time = 34 s (3 s after contact with carbachol), cell length is 237 µm; third panel: time = 35.3 s (4.3 s after carbachol), cell length is 162 µm; fourth panel: cell after contraction was complete, time = 41 s, cell length is 85 µm.

The shortening response was treated as an exponential process, as described in previous work (9). Obviously, the cells do not contract to zero length, but rather their length asymptotically approaches some minimum length, representing the maximum contraction of which they are capable under the chosen conditions [physiological Ca²⁺ concentrations and high (100 µM) agonist concentrations]. Usually the cells contract to one-half or less of their starting length for both adult and preterm cells.

The shortening velocity v can be compared with estimates of unloaded shortening velocity of multicellular preparations expressed in lengths per second. The value of 0.503 s¹ for the cell shown in Fig. 2 is greater than most estimates from multicellular ASM preparations. These results demonstrate that the contractile system in the single cells is well preserved after the isolation procedure.

Figure 3 shows the contraction of a preterm ASM cell in response to carbachol. As shown, Figure 3A is a plot of cell length vs. time. The carbachol-Trypan blue solution reached the preterm cell at 41.5 s, the cell was contracting at 44 s, and the contraction was essentially complete by 50 s. As with the adult cell, the contraction was substantial (final length ~22% of initial length). The shortening velocity of this cell was 0.788 s¹, which is more than the value for the adult cell shown. Figure 3B shows images of the preterm cell as it contracted from a length of 179 µm to a final length of 40 µm.

View larger version (13K): [in this window] [in a new window]   View larger version (162K): [in this window] [in a new window]   Fig. 3.   Contraction of a preterm ASM cell. Same conditions as in Fig. 2. A: plot of cell length as a function of time during shortening. Agonist reached cell at a time corresponding to 41.5 s, denoted by downward arrow. Symbols are as described in Fig. 2 legend. B: photomicrographs of same cell at selected points in contraction. First panel: cell at a time (42.5 s) after carbachol reached it but before it started to contract, cell length is 179 µm; second panel: time = 44 s (2.5 s after contact with carbachol), cell length is 140 µm; third panel: time = 45 s (3.5 s after carbachol), cell length is 90 µm; fourth panel: cell after contraction was complete, time = 58 s, cell length is 40 µm.

Table 2 summarizes the shortening velocities and the amounts shortened by adult and preterm cells. The shortening velocities of the preterm cells were not significantly different from the adult cells. However, there was a significant difference in the amount shortened: the preterm cells shortened a greater fraction of their cell length (68%) than the adult cells did (55%).

Other agonists besides carbachol cause rapid contractions of these cells. In a preliminary study, a sample of 10 adult cells stimulated with 10 mM caffeine in 1.6 mM CaCl₂ exhibited a mean shortening velocity of 0.59 s¹ (range 0.42-0.74 s¹). Furthermore, the shortening velocity with 100 µM serotonin was similar (mean 0.56 s¹; range 0.46-0.65 s¹). Thus the response of the cells and the suitability of the measurement method are not limited to carbachol stimulation. Further experiments need to be done to fully characterize the response of ASM cells to these and other agonists.

	DISCUSSION

We believe that a smooth muscle cell preparation can produce the highest quality results only if it can meet the following criteria: 1) the cells should be relaxed and elongated so that substantial contraction can be demonstrated; 2) they should be Ca²⁺ tolerant at physiological temperatures, meaning that they should maintain their elongated state when transferred from Ca²⁺-free solutions to ones with physiological (mM) Ca²⁺ concentrations; 3) they should exclude Trypan blue; 4) they should contract within seconds of being exposed to contractile agonists; and 5) the contractions should be substantial, meaning one-half of the cell's starting length. One of the reasons that studies on the Bufo cells have produced so many important results is that the cells used were of very high quality and met these standards. Using the method described herein, we have been able to isolate elongated, relaxed, single smooth muscle cells from both adult and preterm sheep tracheae. These cells remain relaxed in millimolar concentrations of Ca²⁺ and contract substantially and rapidly when exposed to cholinergic agonists.

A common problem if smooth muscle cell preparations are to be used for contractile studies is that the isolated cells are extremely short, possibly because they have already contracted during the isolation procedure. The conclusion that cells are precontracted is only valid if it can be later demonstrated that elongated cells can be prepared from the species and tissue in question. Cells in other species and tissues might just be intrinsically shorter, or differences in the amount and nature of cytoskeletal proteins expressed might be responsible. Maximum shortening of 70% of the starting length can be seen in healthy, elongated ASM cells. When our preparations deteriorate, ~8 h postdigestion, they commonly lose their Ca²⁺ tolerance and shorten, and these damaged cells are then only capable of 10-20% shortening when agonists are added.

ASM cells suitable for patch-clamp studies of ion currents have been isolated and used successfully in a number of studies (16, 19, 20, 32, 34). In one case, the electrophysiological studies were combined with simultaneous measurement of contraction, a very powerful approach (19). These canine tracheal cells had a mean length of 119 µm and rapidly shortened 20-40% of their initial length when exposed to acetylcholine. Additionally, muscarinic receptors were studied in isolated bovine, canine, and porcine ASM cells (19, 21, 23, 34).

After contraction in response to 100 µM carbachol, the ASM cells usually do not elongate, even though the agonist-containing solution is washed out of the perfusion dish. This behavior is typical of vascular smooth muscle cells as well. Alexander and Cheung (1) found that, although their arterial cells did not elongate after washout of the stimulus, the free Ca²⁺ concentration (estimated by using fura-2) did return to the initial resting value. Thus the failure of a smooth muscle cell to elongate after contraction is not necessarily an indication of abnormal Ca²⁺ handling. When our ASM cells contract in response to lower concentrations of carbachol, a variable amount of elongation can be observed after the agonist is washed out (not shown). In the best example, a 187-µm cell briefly stimulated with 10 µM carbachol shortened to a length of 83 µm and then elongated to 123 µm after the carbachol was washed out, a 48% increase in length.

The only real disadvantage of our isolation method is the relatively low yield of cells. We estimate that it might be possible to isolate ~5 × 10⁵ cells from the tracheal smooth muscle in a 4-cm segment of sheep trachea, if it were completely teased into fine strips. Our method is fundamentally different from most cell-preparation methods in that papain is only used to soften the tissue; the actual release of cells takes place when fine strips are dissected from the digested tissue in a Ca²⁺-free solution. Most other methods use crude collagenase mixtures, and many investigators now also add papain to these collagenase mixtures. In those preparative methods, relatively large numbers of cells are liberated by the combined action of proteolytic enzymes and agitation, and the teasing of fine strips from the digested tissue is not required.

Researchers who use collagenase mixtures for cell dissociation periodically have difficulties caused by the variability in batches of collagenase. An advantage of papain is that it is a crystalline enzyme, and we have never purchased an unsatisfactory lot. Another advantage of papain is that it does not require Ca²⁺ for enzyme activation, so Ca²⁺-free conditions can be used if desired.

We routinely used high concentrations (100 µM) of carbachol to cause maximal contractions, but lower concentrations (10 and 1 µM) also cause contractions. We have not yet determined the responsiveness to threshold doses of cholinergic agonists, nor have we systematically investigated other potential contractile agonists. Many smooth muscle cell preparations are more sensitive to cholinergic agonists than are the intact tissue from which they were derived, but this is not always the case. For example, toad stomach cells are more sensitive to acetylcholine than the muscle strips are, yet the isolated cells are less sensitive to carbachol than are the muscle strips, needing nearly 10⁴ M carbachol for a maximum cell response (12). Adult sheep tracheal strips require high concentrations of acetylcholine (10³ M) to achieve maximum force (25). Preliminary studies showed that KCl, caffeine, and serotonin caused isolated adult cells to contract. The preliminary results with caffeine are important because they demonstrate that the rapid shortening response is not restricted to cholinergic agonists. We presume that carbachol contracts these cells through inositol trisphosphate-mediated release of Ca²⁺ from the sarcoplasmic reticulum. Caffeine is believed to have a simpler mode of action, directly causing Ca²⁺ release from the sarcoplasmic reticulum, although activation of plasma membrane cation channels may also take place (15). The similarity of shortening velocities with carbachol, caffeine, and serotonin is reassuring and suggests that the contractions we observed may be the maximum of which these cells are capable under these conditions. Further studies need to be done to determine whether there are differences between the adult and neonatal cells with respect to functional receptors expressed and their sensitivity.

Shortening velocities of the isolated airway cells were greater than most reported values of maximum shortening velocity in multicellular ASM tissues (35, 36) and are about three times faster than the shortening velocities of isolated arterial smooth muscle cells (9). The more rapid shortening of isolated cells compared with intact multicellular tissues may be due to the elimination of tissue elements that reduce shortening velocity in the multicellular preparations.

The technique described here for delivering agonist to cells by switching to a slightly denser agonist solution has advantages compared with other approaches. It is less complicated than trying to rapidly move a multiwell apparatus underneath a cell held by a microtool. Another approach is to use a device that sprays picoliter amounts of agonist solution onto a cell in a flowing solution. With the "spray" method, the agonist concentrations are not uniform around the cell, nor is the agonist concentration known. The presence of 1% sucrose in the agonist solution increases osmolarity (~35 mosM). If this is unacceptable, solutes such as Percoll can increase the density with less of an increase in osmolarity.

Cell length will be underestimated when a cell is not lying flat on the floor of the chamber. This source of error tends to be self-limiting because the part of the cell that is not lying flat will be out of focus. If the end of the cell cannot be identified in the image, length measurements are impossible. We found that, with our apparatus, focusing the ×40 (0.60 numerical aperture) objective up or down more than 20 µm made the image of the end of a cell unacceptably blurry. In our experiments, we took this distance as the maximum distance out of the focal plane that the end of a cell would be.

This allowed us to estimate the error in cell lengths by analyzing the geometry of cells of different lengths resting with one end remaining 20 µm above the plane of the chamber floor. As expected, these errors are negligible for very long cells and are increasingly important for shorter cells. Before contraction, a 200-µm-long cell would appear to be 199 µm long (a 0.5% error). When it shortens to a length of 100 µm, the cell would appear to be 98 µm long (a 2% error). Finally, when its true length reaches 50 µm, it would appear to be 46 µm long (an 8% error). This would lead us to state that the cell shortened to 23% of its starting length, rather than the true 25% of starting length, so we would slightly overestimate the amount the cell has shortened. These minor errors do not cause serious problems with the accuracy of our velocity measurements, because the shortening velocity we calculate is dominated by the earlier time points when the cell is longer and can be measured more accurately. Taking into account the various sources of possible error, we estimate conservatively that the length measurements of unstimulated cells are accurate to 5% of the true value.

This method for isolating cells, delivering agonist, and analyzing shortening velocity is generally applicable to smooth muscle tissues from various species, and we have successfully applied it to sheep pulmonary arteries, guinea pig stomach, and rat colon (unpublished observations). We have shown here that the method works with animals of various stages of development, from preterm lambs to adult sheep. The apparatus for recording contractions described here is relatively simple and does not use expensive equipment. This preparative method, along with the analytic method of measuring shortening velocity, can help researchers studying various smooth muscles to make important measurements of the most fundamental property of muscle cells, namely, contraction.