cwcakfm(1ea2~13, m-00 OLcolJmlnQrcupuKLtd1ee2

Subcellular distribution of cytosolic Ca*+ in isolated rat hepatocyte couplets: Evaluation using confocal microscopy

M.H. NATHANSON and 9.5. BURGSTAHLER

Deparlments of Medicine and Cell Biology and the Liver Center, Yale University School of Medicine, New Haven, Connecticut, USA

Abstract - Ca2+ agonists induce Ca2’ waves and other non-uniform Ca2’ patterns in the cytosol of epithettat cells. To define subcellular Ca2+ transients in the cytosoi of hepatocytes we examined Flue-3-loaded isolated rat hepatocyte couplets using confocai microscopy. Optical sections of less than 1 micron in thickness were observed in couplets, and fluorescence from cytosoiic Ca2+ signals was readily distinguished from nuclear, mitochondriai, and iysosomai fluorescence. The nature of the noncytosolic components of the fluorescent images was verified by double iabeiling with the mitochondrial dye DIOCe(3) and with the iysosomai marker acridine orange. Using the line scanning mode of confocai microscopy, measurements of cytosoiic Ca2+ were made with a frequency of up to 25CI Hz and without significant bleaching. it was found that phenyiephrine-Induced Ca2’ signals generally began at the basal pole of the hepatocytes, then spread to the oanaliculus at average speeds of 80 *s. These findings demonstrate the utikty of confocal tine scanning microscopy for detecting rapid changes in the subcellular distribu$ion of cytosoiic Ca2+ in hepatocyte couplets, and suggest that phenyiephrine-induced Ca2+ waves radiate in a basal-to-apical direction in this ceil type.

Cytosolic Ca2’ (Ca?) is a second messenger for a number of important hepatocyte functions, ranging from glycogenoiysis [l] to canalicular contraction [2] to bile secretion 131. A variety of hormones and drugs lead to increased Cai2’ in hepatocytes, and this increase is often mediated by a rise in inositol t&phosphate (IP3) [ll. lP3-mediated Cai2’ IiSeS appear to begin in a single intracellular locus [4], then spread at speeds ranging from 20-25 [41 to over 200 @s 151; similar Ca?’ waves have been described in other epithelial cell types L&8]. Since it is not yet known whether Cai2* waves are

specifically involved in the regulation of hepatocyte function, it would be important to measure the subcellular distribution of Cai2’ in hepatocytes.

It has been difficult to adequately define the subceilular distribution of hormone-indnced Cai2+ signals in isolated hepatocytes; Ca12’ waves typically traverse these cells in less than 1 s [5] and release Of Caged IP3 IZSUltS itI localized Cai2+ increases within 50 ms 191, so that Cai2+ transients occur at or exceed the limits of detection by many digital epifluorescence imaging systems. In addition, epithelial cell preparations in which

89

90 CELL CALCIUM

polarity has been maintained are often too thick or dense for de&&d subcellular resolution of Cat+ dyes by digital epifluorescence imaging. However, studies of another epithelial cell type, pancreatic exocrine cells [8], suggest that the significance of Cai2+-waves in epithelia may only be apparent in polarized cell preparations. To circumvent these problems, we used confocal line scanning microscopy [lO-121 to examine isolated rat he atocyte couplets [ 13, 141 loaded with the

5+ Ca -sensitive dye Flu+3 [15].

Materials and Methods

Animals and materials

Male Sprague-Dawley rats (180-250 g; Camm Research Lab Animals, Wayne, NJ, USA) were maintained on Purina rodent chow under a constant light cycle and used for all experiments. Phenylephrine and arg*-vasopressin were obtained from Sigma Chemical Company (St Louis, MO, USA), 2,5-di-tert-butylhydroquinone (tBuBHQ) was obtained from Aldrich Chemical Company (Milwaukee, WI, USA), ionomycin was obtained from Calbiochem (San Diego, CA, USA) and Flue-UAM, acridine orange (high purity) and 3.3’~dihexyloxacarbocyanine iodide (DiOC6(3)) were obtained from Molecular Probes (Pitchford, OR, USA). All other chemicals were of the highest quality commercially available.

Preparation of isolated hepatocytes and hepatocyte couplets

Isolated rat hepatocytes and hepatocyte couplets were prepared in the Hepatocyte Isolation Core Facility of the Yale Liver Center as described previously [13, 141. Briefly, rat livers were perfused with Hanks A then Hanks B medium containing 0.05% collagenase (Boehringer Mannheim Biochemicals, Indianapolis, IN, USA) and 0.8 U trypsin inhibitor (Sigma Chemical)/U tryptic activity. Livers were then excised, mine passed through serial nylon mesh filters, and the resultant cells washed. Cells were suspended at a concentration of 5 x 10’ cells/ml in Liebovitx L-15

medium (GIBCO, Grand Islaud, NY, USA) containing 50 U penicillin and 50 pg streptomycin per ml, and plated onto glass coverslips. Cell viability by trypan blue exclusion exceeded 85%.

Confocal jluorescence imaging of cytosolic calcium

Isolated rat hepatocytes and hepatocyte couplets were prepared and plated onto glass coverslips as described above, incubated at 37°C for 26 h, then loaded with the Ca2+-sensitive fluorescent dye Fluo-3/AM (6 pM) [15] in L-15 medium containing 10% fetal calf serum for 20 min at either 15’C, 22’C (room temperature), or 37°C. The coverslips containing these cells were transferred to a chamber on the stage of a Zeiss Axiovert microscope and the cells were perfused at 37°C with a HEPRS-buffered solution and observed using a BioRad MRC-600 confocal imaging system An argon laser was used to excite the dye at 488 nm and emission signals were collected using a 515 nm barrier filter. Optical sections between 0.5-1.0 pm in thickness were obtained in the hepatocytes; in short-term culture these cells are spherical with a diameter of 20-25 i.un. Neither autofluorescence nor other background signals were detectable at the machine settings (i.e. apertnre, gain, and black level) that were used. For comparison, some of the coverslips containing Flue-3-loaded hepatocytes were transferred to a perfusion chamber maintained at 37°C on the stage of a Zeiss IM35 microscope and the cells were observed at the same magnification using a Hamamatsu C2400 SIT camera (Hamamatsu Photonics K.K., Hamamatsu City, Japan) and digitized. Noise was reduced in these images by on-line averaging using an Rex Series 151 image processor.

Confocal microscopic double labelling studies

To define intracellular regions of decreased Flue-3 fluorescence, some of the hepatocytes were simultaneously labelled with Fluo-3 (as described above) and DiOCa(3). DiOC6(3) is a lipophilic dye that preferentially partitions into membranes of mitochondria and endoplasmic reticulum [161. The dye was dissolved in 100% ethanol (0.2 mg/ml), and Fluo-3-loaded hepatocytes were incubated at 37’C

Ca2+i IN HEPATOCYTE COUPLETS BY CONFOCAL h4KROXOPY 91

for 5 min in rmxlium with a final concentration of 0.5 ug DiOCe(3)/1nl. The excitation and emission peaks of DiOCa(3) are 483 and 499 nm, respectively, and emission fluorescence is reduced to -13% of peak intensity at 560 nm. In contmst, Flue-3 fluorescence is decreased by only -50% relative to peak intensity at this wavelength. Therefore, double-label&l couplets were excited at 488 nm and separate emission signals were collected simultaneously at 540 nm (bandpass) and > 600 run (long pass) by confocal microscopy. Using this approach, the signal from Di0c6(3) wa!? the

predominant one at the lower emission wavelength, while a signal pmdominantly from Flu@-3 was detected simultaneously at the higher wavelength range.

Couplets were loaded with acridine orange (AO) to define intracellular regions of increased Fluo-3 fluorescence. The dye was dissolved in 100% ethanol (2 mg/ml) and cells were incubated with 25 p.M A0 at 37°C for 20 min. The intracellular distribution of the dye was examined by confocal microscopy and compared to the distribution of Flue-3. Double labelling studies with Fluo-3 and A0 were not possible since A0 quenches fluorescence of other nearby fluorophores [17]. However, the weak base chloroquine accumulates in and causes swelling of the acidic compartments in which A0 collects [17], so that the effects of chloroquine on the distribution of compartment- alized Flue-3 fluorescence could be compared to its effects on A0 fluorescence. Cells loaded with either A0 or Fluo-3 as described above were perfused at 37°C in the chamber on the stage of the confocal microscope. Images of isolated cells and couplets were recorded under basal conditions, then chloroquine (200 pM) was added to the perfusate and confocal images were obtained serially over 10 min.

Confocal line scanning measurements Of Cai2+

Single cells and couplets were stimulated under one of the conditions described below and the resulting Cal=+ signals were detected by confocal line scanning microscopy. In the line scanning mode of confocal microscopy, fluorescence is determined at each point along a single line across the image,

rather than at each point across the entire image, which allows fluomscence data to be collected as frequently as every 4 ms 15,101.

Couplets were scanned 512 times along the scan line, at fixed time intervals between &al scans. To assess basal rate of loss of Flue-3 fluorescence along the scan line over time (due to bleaching, dye leakage or secretion, and ongoing comp;atmental- ization), some couplets were scanned under control (unstimulated) conditions. To compare patterns of Ip3 and non-W mediated cai2’ responses, c?tlets were stimulated with either phenyleptuiue (10 vasopressin (1 O-*

M), M), ionomycin (lOA M), or

tBuBHQ (2.5 x lo-’ M). Line scans of couplets under control conditions and stimulated with each of these agents were collected with a scan interval of 0.5 s (2 Hz), which is a typical sampliug interval in nonconfocal imaging systems, and additional scans of couplets stimulated with phenylephrlne or vasopressin were collected with a scan interval of 4 ms (250 Hz). Fluo-3 fluorescence in response to each of these stimuli was recorded by this method of confocal line scanning microscopy, and the gain and black level were adjusted to allow the full dynamic range of fluorescence response to be detected while eliminating all background fluores- cence. The recorded images were low-pass filtered and scaled to peak brightness using an Itex Series 151 image processor, and the change in fluorescence over time at each point along the scan line was then determined from the recorded image.

Results and Discussions

Representative non-confocal (epifluorescent) and confocal images of Fluo-3-loaded couplets are shown in Figure 1A and lB, respectively. In the nonconfocal image (Fig. lA), fluorescence appears to be uniform throughout much of the cytosol and increased in the pericanalicular region and within the canaliculus. In the confocal image (Fig. lB), in contrast, there is an intricate pattern of intracellular fluorescence. The marked difference in appearance between the epifluorescent image (Fig. 1A) and the confocal image (Fig. 1B) reflects that confocal microscopy provides improved axial resolution and removes out-of-focus fluorescence [l 1, 12, 181.

CELL CALCIUM

Ng. 1 Comparison of epifiuomscent and confocal fluorescent images of isolated rat hepatocyte couplets loaded with the Ca2+-sensitive

dye Fluo-3 A. Epifluomscent image. Note the limited spatial resolution; the entire canalicuhu region appeals to have increased Ca2’ relative to the

remainder of the cytosol, but cytosolic Ca” can not be further localized. Objective: Zeiss Neofluar 63x, 1.25 NA

B. Confocal fluorescent image. Here it is apparent that there are punctate increases in fluorescence within the pericanalicular cytosol,

and the canalicular lumen is tv%ually devoid of fluorescence. Elsewhere throughout the cytosol then2 am numerous small regions in which fluorescence is decnased. The nucleus is also distinguished by a focal decrease in fluorescence. Objective in this and subsequent

_ confocal images: Zeiss Plan-Apochromat 63x, 1.40 NA

Several types of fluorescent labelling studies were undertaken to define the basis of the subcellular fluorescence pattern in Figure 1B. Double-labelling studies with DiOCs(3) indicate that cytosolic regions not labelled by Fluo-3 are occupied by mitochondria (Fig. 2). The intracellular distribution of this organelle as determined by these confocal images is similar to mitochondrial localization within hepatocyte couplets as described by electron microscopy [191. The concentration of DiOG(3) used here has also been used to label endoplasmic reticulum [ 161, although this organelle was poorly visualized in these studies. mectron microscopic studies have indicated that the typical distribution of endoplasmic reticulum in isolated rat hepatocyte couplets is predominantly in the basal region of the cytosol and the subplasmalemmal region of the basal membrane [19]. The current work augments information from electron micro- scopic studies in that it allows simultaneous

visualization of Cat+ pools and mitochondria, and allows these various intracellular regions to be visualized in non-fixed cells.

Two types of studies were conducted to define subcellular regions of increased fluorescence. Punctate pericanalicular regions of increased Fluo-3 fluorescence (Fig. lB, Fig. 2B) were distributed very similarly to regions of increased A0 fluor- escence (Fig. 3). Addition of chloroquine resulted in coalescence and expansion of these regions within 5-10 min in Fhro-3-loaded couplets (Fig. 4). Chloroquine induces this morphological change in the distribution of A0 in a number of cell types 1171, including hepatocyte couplets (data not shown). Together, these findings suggest that regions of increased Fluo-3 fluorescence in hepatocytes are identical to the acidic intracellular compartments into which A0 accumulates [17, 201. These results also indicate that Fluo-3 compart- mentalizes in a manner similar to Fura- [17].

Ca*+i IN HEPATOCYTE COUPLETS BY CONFOCAL MICROSCOPY

Mg. 2 Double labelling of an isolated rat hepatocyte couplet with DiOCe(3) (left) and Fluo-3 (right), as detected by conf

microscopy. The image on the left illustrates the mhochondria and, to a lesser extent, the endoplasmic reticulum, which

fluorescently labelled with DiOa(3). The image on the right illustrates the distribution of the Ca%-sensitive dye Fhto-3, pdomin2

witbin the cytosol but with some pericanalicular compartmentalization. The regions of increased fluot~cence of Flue-3 am coincident with DiOCdf)-labelled organelles, the latter of which are the Ca’+-storing organelles in the hepatocyte. This demonsa

that these organelles are not the pools that have been fluorescently labelled for Ca’+ with Flue-3

Fig. 3 Confocal image of an isolated rat hepatocyte couplet loaded with acridine orange. Note pericanalicular regions of increased fluorescence (arrows). These represent acidic vesicles, including

lysosomes

Fig. 4 Confocal image of a Fluo-3-loaded isolated rat hepatocyte couplet a&r exposure to chloroquine (200 pM) for 10 mm. Pericanalicular regions of increased (‘compartmentalized’) fluorescence have coalesced and expanded (compare to Fig. 15)‘

an effect of chlomquine on acidic compartments. Cain and black level in this image were adjusted to maximiae contrast between compartmentalized and cytosolic fluorescence

94 CELL CALCIUM

Thus, confocal microscopy can be used to distinguish cytosol from mitochondria, nucleus, and acidic (lysosomal and late endosomal) vesicles in Flue-3-loaded hepatocytes.

Compartmentalization of Ca2+-sensitive dyes into lysosomes has been investigated extensively and is sometimes dependent on the temperature at which cells are loaded. using Flue-3, pericanalicular compartmentalization was seen in over half of hepatocyte couplets loaded at either 15°C. 22’C (room temperature), or 37°C. This suggests that some of de-ester&d Fkk3 becomes trapped within the endosomal/lysosomal pathway regardless of loadin temperatme.

2F colnpart-

mentalization of Ca -sensitive dyes has been a problem for epifluorescence measurements of Cat+ signals because such compartments introduce contributions to cellular fluorescence which do not represent cytosolic Ca2+ [17]. In confocal images of several representative hepatocytes, the fraction of cross-sectional area occupied by compartmentalized dye was l-246, while the cytosolic fraction of FM-3 was 50-6046 and was readily distinguished from regions of compartmentalized dye. In addition, cytosolic fluorescence as detected by confocal microscopy is not confounded by fluorescence from compartmentalized dye in other focal planes. Unlike epifluorescence microscopy, confocal microscopy thus allows fluorescence from cytosolic Ca2’ to be readily distinguished from compartmentalized fluorescence, so that compart- mentalization of dye does not complicate measurements of cytosolic Ca2’ signals. Moreover,

confocal microscopy can be used to make repetitive measurements of such Ca? signals at specific locations within the cytosol.

A major problem with the initial Ca2+-sensitive dye Quin-2 was inability to account for decreased fluorescence due to photobleaching or dye leakage [213. This problem was largely obviated when the ratio dyes Fura- and Indwl were introduced [21, 221. The Ca2’ dye Flue-3 is distingnished fYom each of these previous dyes because it can be excited using an argon laser, the light source in many confocal microscopes [5, 10, 151. Like Quin-2, however, Flue-3 has single peaks for both excitation and emission, so it can not be used for ratio imaging [15]. DecIeases in Fkk-3 fluorescence, presumably due to photobleaching, are rapid in serial confocal scans of nonstimulated isolated hepatocyte couplets (data not shown). However, fluorescence exposure is rednced by a factor of 512 when cells are imaged by confocal line scanning microscopy, since only one of the 512 lines of the digitized image is collected. In addition, fluorescence recovery after photobleaching may occur along the scan line, due to diffusion of cytosolic Fluo-3. A non-stimulated couplet and the corresponding line scan is shown in Figure 5A and SB, respectively. To quantitate fluorescence loss due to this line scan, the change in fluorescence over time was examined at four representative points along the scan line. Statistics which characterize the fluorescence measun?ments at each of these four points are summan ‘zed in Table 1. Linear regression was performed for each of the four time cnrves, and the slopes of the regression lines wefe taken to indicate the rates of bleaching at each of the corresponding points. The average rate of bleaching was 0.01% per scan, or 3.8% per 512 scans; similar figures were obtained regardless of the time interval between scans. These results

Table 1 Photobleaching of Pluo-3 using confoeal line scanning microscopy

FlUOWSC~lUX CMflCieN B&aching COlV&tlO?t

Intcnsily’ of varhtion (%) rat2 coqpckmt

129 f 12 9.3 0.013 0.16

138 f 12 8.7 0.004 0.05

182 f 13 7.1 0.024 0.24

210f 11 5.2 0.008 0.09

Ca”i IN HEPATOCYTE COUPLETS BY CONFOCAL MICROgCOPY 95

Fig. 5 Loss of Flue-3 fluorescence during confocal line scanning

micIoscopy

A. Confocal image of an isolated rat hepatccyte couplet loaded

with Flue-3. Horizontal hue moss the field is the line along which the indicated couplet was subsequently scauned

B. Confocal line scan of the couplet in (A), performed along the line indicated in that figure. Bach horiaomal line in this figure

represents the flmmxcence along each corresponding point of the horizontal line indicated in (A). Each of these lines was obtained

after the line immediately above it, for a total of 512 consecutive

lines. Each vertical line in this &gum thus corresponds to a

separate point on the line in (A) and shows the change in

fluorescence over time at that point The fluorescence intensity

does not decrease along any vertical line in this figure, which

indicates that no measurable loss of fluorescence occurs over time

at any point along the (horizontal) line that is scanned

indicate that loss of fluorescence signal is negligible after 512 consecutive confocal line scans. Frequency histograms for each of the four time curves were also examined. Each histogram was normally distributed about its mean fluorescence value (P > 0.9, chi-square goodness-of-fit test [23]), with coefficients of variation ranging from 5 to 9%. Thus, measurement errors of Flu*3 fluorescence at each point along a line scan are small and follow a normal (Gaussian) probability distribution [23]. Together, these observations demonstrate that Flu03 fluorescence can be measured up to 512 times along a single line in a cell or couplet using confocal line scanning microscopy, and that the measurements will be highly reproducible and negligibly corrupted by loss of fluorescence. Thus, quantitative studies of changes in Ca12’ can be performed in this cell system with confocal microscopy, provided that the line scanning

approach is used. The demonstratioo that line sWlnirlg measurement errors are uoimally distrlbutedfurtberindicatesthat,usingdataaquired by this approach, parameters from mathematical models of Ca?+ kinetics can be estimated using staudard statistical parameter e&nation techniques [241.

Confocal line scans of couplets stimulated with different agonists were obtained at two different rates, 2 Hz (500 ms between scans) and 250 Hz (4 ms between scans). Couplets stimulated with tBuBHQ, ionomycin, phenylephrlne, or vasopressin were scanned at 2 Hz, and couplets stimulated with phenylephrine or vasopressin were also scanned at 250 Hz. Results from representative couplets stimulated with phenylephrine or tBuBHQ are shown in Figure 6. Each of these line scam was obtained over 256 s with a sampling interval of 0.5 s, to illustrate Ca2+ signaling as detected at a rate typical of nonconfocal systems. Increases in Ca? in couplets stimulated with tBuBHQ (Fig. 6) or ionomycin (not shown) occurred progressively over seconds; neither of these agents induce a rise in Ca12+ which is mediated by IP3 [25, 261. In contrast, couplets stimulated with phenylephrine (Fig. 6) or vas;fressin (not shown) revealed an increase in Cai which was instautaneous and uniform throughout the couplet, as would be expected since Cai2 signals tiom both of these agents are mediated via IP3 [l]. Couplets stimulated with phenylephrine were also examined by line scanning with a sampling interval of 4 ma over a total of 2 s (Fig. 7). The Ca12+ rise appears nonuniform in this couplet, beginning in a single intracellular locus and spreading at a rate of 88 elm/s, This observation agrees well with the previously reported pattern of Cal’+ signaling in IRHC detected by confocal line scannin microscopy, in which vasopressin-induced Cal 2g

waves travelled across couplets at average speeds of over 100 pm/s [5]. Phenylephrine- and vasopressin- induced Cai2+ waves have also been detected in isolated hepatocytes by epifluorescent ratio imaging of Fura- [41. In contrast to the current fmdings, Ca12+ waves in that study were estimated to travel at only 21-23 pm/s [41. Those results were obtained near the limits of temporal resolution of the imaging system used (1.1 Hz), so that Ca12’ waves were

0 20 40 60 60 100 0 20 40 60 80 100

time (WC) timr (WC)

Fig. 6 Confocal line scanning mic~~~py of isolated rat hepatocyte couplets stimulated with tBuBHQ (25 @I) or pknyleph&e (10

WI A. Confocal image of Flue-3-loaded isolated hepatocytes and hepatocyte couplets immediately prior to stimulation with tBuBHQ.

Horizontal line across the field is the line along which the indicated couplet was subsequently scanned

B.Confocal line scan of the couplet in (A). Infusion of tBuBHQ began at the indicated time, and the subseqmnt gradual increase in

fluorescence indicates the induced rise in Cap. Scans were obtained every 0.5 s, over a total of 256 s

C. Change in fluomscencc over time at a single point along the scan line in (A). Infusion of tBuBHQ begins at t - 0 s, and after a brief latency period, Flue-3 fluorescence increases over the next -8 s to its maximum value. This relatively gradual rise in Cat2’ occurred

simultaneously throughout the couplet (Fig. 6B)

D. Confocal image of Flue-3-loaded isolated hepatocytes and hepatocyte couplets immediately prior to stimulation with phenylephrine.

Horizontal line across the field is the line along which the indicated couplet was subsequently scanned E. Confocal line scan of the couplet in (D). Infusion of phenylephrine began at the indicated time, and the subsequent instantaneous increase in fluorescence indicates the induced rise in Cap. Scans were obtained every 0.5 s, over a total of 256 s F. Change in fluorescence over time at a single point along the scan line in (D). Inkion of phenylephrine begins at t - 0 s, and at& a brief latency period, Fluo-3 fluorescence appears to increase nearly instantaneously to its maximum. This rise in Cap occurred simultaneously throughout the couplet (Fig. 6E)

detected using 2-4 consecutive epifluorescent emanate from a single intxacellular source [4], the images rather than using 12-50 consecutive previous observations were made using single confocal images, as in the current study. isolated hepatocytes, which have lost their

Although the current work confirms the previous finding that phenylephrine-induced Cai2’

morphological and functional polarity. In this study, waves phenylephrine-induced Ca?’ signals were examined

Ca’+i IN HEPATOCYTB COUPLETS BY CONFOCAL MICROSCOPY 97

c 3 250 2 9 z 225 x ‘B - 200 21 .Z

E 175 % .-L

:: 150 s M ? 125

ci lOOi E

0.0 0.2 0.4 0.6 0.8 1.0 1.2

time (aec)

Fig. 7 Confocal line rxxnning microscopy of an isolated rat

hepatocyte couplet stimulated with phenylephrine ( 10V5 M)

A. Confocal image of Fhr+3-loaded hepatocytes immediately

prior to stimulation with phenylephrine. Horizontal line across the

field is the line along which the indicated couplet was

subsequently scanned

B. Confocal line scan of the couplet in (A). Scans were obtained

every 4 ms for a total of 2.048 s. The induced rise in Ca?’ begins

at the basal pole of the right cell and advances across the couplet

C. Change in fluorescence over time at two points, 14 pm apart,

along the scan line in (A). The increase in fluomscenoe at one of

the points (dashed line) precedes the increase at the other point

(solid line) by 160 ms, so that the Cai *+ wave navels between the

two points at a speed of 88 pm/s

in 7 hepatocyte couplets and 1 triplet (II = 17 cells), with retained polarity. Cai2+ waves could be identified in 12 of these cells, and the Cai2’ * mcrease

appeared to occur simultaneously throughout the cytosol in the remaining 5 cells. The average speed of the Cai2+ waves was 80 f 34 pm!s (range 38-135 un&). In one of the couplets and in the triplet, the

Cai2+ wave began at the basal Role of one of the cells and spread at a constant rate across that cell and the adjacent cell(s). The Ca? wave began at the basal Role of 4 of the other hepatocytes and at the apical pole of the remaining 3 cells. Thus, Cai2+ signals began more frequently at the basal pole, then spread to the canaliculus. Internal organization of Ca? signals has also been described in hepatocyte couplets stimulated with V66OpreSsin, where Cai2+

signals ate highly co-ordinated between cells [5]. Although apical-to-basal Ca?’ waves regulate directed Cl- secretion in pancreatic epithelium [8], the functional significance of basal-to-apical Cay waves in hepatocytes is not clear. Recent studies have shown that increased Cai2+ phosphorylates pericanalicular myosin light chain, which leads to contraction of the canalicular lumen 1271, and that canalicular contractions are co-ordinated among adjacent hepatocytes to direct the flow of bile [28]. From these observations, it may be hypothesized that apically directed Ca? waves regulate canalicular motility, but this and other roles of hepatocyte Ca? waves remain to be established.

Conclusions

Confocal line scanning microscopy is a useful method for assessing subcellular patterns of Cai2+ signaling in isolated rat hepatocyte couplets. With this approach, it was observed that Ca?’ waves induced by phenylephrine usually spread from the basal pole of hepatocytes to the canaliculus, crossing the cells within one-quarter of a second. Loss of fluorescence over time was negligible using this method, so that certain quantitative studies may be performed with Fluo-3 even though this dye can not be ratio imaged. Further studies will be needed to determine whether specific hepatocyte functions are regulated by intra- and intercellular Cai2’ waves.

Acknowledgements

We thank Johanna M. KleinRobbenhaar for preparation of isolated hepatocytes and Albert Mennone for expert photographic assistance. This work was suppaittd by the Morphology Core and Hepatocyte Isolation Core Facilities

98 CEIL CALCIUM

of the Yale Liver Center (NIH P30 DK34989). a Liver Center Pilot Project Award (MHN), a Clinician-Scientist

Award from the American Heart Association (MHN), a Postdoctoral Research Fellowship Award from the American Liver Foundation (MI-IN), and a grant fkom the Mather Foundation.

15.

16.

Kao JPY. Hamotunian AT. Tsien RY. (1989) Photochemically generated cytoaolic calcium pulses and their detection by flue-3. J. Biol. Chem., 264.8179-8184. Terasaki M. Song JD. Wong JR Weiss MJ. Chen LB. (1984) Localization of endoplasmic reticulum in living and glutaraldehyde-fixed cells with fluorescent dyes. Cell, 38, 101-108.

17.

References 18.

1. Williamson JR. Cooper RH. Joseph SK. Thomas AP. (1985) Inositcl trisphosphate and diacylglyccrol as intracellular second messenger in liver. Am. J. Physiol., 248, C203C216.

Malgamli A. Milani D. Meldoleai J. Poaxan T. (1987) Fura- measurement of cytosolic free Ca% in monolayeta and suspensions of various types of animal cells. J. Cell. Biol., 105,2145-2155. NiggliE. Le&rer WJ. (1990) Real-time amfocal micmscopy and calcium measummenta inheartmusclecells: towards the development of a fluorescence microscope with high temporal and spatial resolution. Cell Calcium, 11, 121-130.

2. Watanabe S. Smith CR. Phillips MJ. (1985) Coordination of the contractile activity of bile canaliculi: evidence from calcium microinjection of triplet hepatocytes. Lab. Invest., 53,215219.

19.

20. 3. Nathanson MH. Gautam A. Bruck R. Bales C. Boyer JL.

(1992) Effects of Ca*+ agonists on cytosolic Ca2+ in isolated hepatocytes and on bile secretion in the isolated perfused rat liver. Hepatology, 15, In pnss.

4. Rooney TA. Sass EJ. Thomas AP. (1990) Agonist-induced cytosohc calcium oscillations originate from a specific locus in single hepatccytes. J. Biol. Chem., 265, 10792107%.

5. Nathanson MH. Burgstahler AD. (1992) &ordination of hormone-induced calcium signals in isolated rat hepatocyte couplets: demonstration with confocal micmscopy. Mol. Biol. Cell, 3, 113-121.

21.

22.

6. Jacob R. (1990) Imaging cytoplasmic free calcium in histamine stimulated endothelial cells and in hlet-Leu-Phe stimulated neutmphils. Cell Calcium, 11.241-249.

7. Sanderson MJ. Charles AC. Dirksen ER. (1990) Mechanical stimulation and intercellular comlnunication increases intracellular Ca2+ in epithelial cells. Cell Regul., 1, 585-596.

23.

24.

25.

Gautam A. Ng OC. Boyer JL. (1987) Isolated xat hepatocyte couplets in short-term culture: structural characteristics and plasma membrane reorganization. Hepatology, 7,216223. Lake JR. Van Dyke RW. Scharachmidt BF. (1987) Acidic vesicles in cultured rat hepatocytee: identification and chamcterixation of their relationship to lysosomea and other storage vesicles. Gasttoenterology, 92.1251-1261. Cobbold PH. Rink TJ. (1987) Fluomscenos and biohnninescence measurement of cytoplasmic free calcium. B&hem. J., 248,313-328. Grynkiewicx G. Poenie M. Tsien RY. (1985) A new generation of Cazf indicators with greatly improved fluorescence properties. J. Biol. Chem., 260,3440-3450. Afifi AA. Azen SP. (1979) Statistical Analysis: A Computer Oriented Approach. New York, Academic Press. Beck JV. Arnold KJ. (1977) Pammeter Estimation in Engineering and Science. New York, John Wiley C Sons. Kass GEN. Duddy SK. Moore GA. Orrenius S. (1989) 2,5-Di-(terr-butyl)-1.4~benxohydroquinone rapidly elevates cytosolic Ca” concentration by mobilizing the inositol 1,4,5-trisphosphate-sensitive Ca2’ pool. J. Biol. Chem., 264, 15192-15198.

8. Kasai H. Augustine GJ. (1990) Cytosolic Ca2’ gradients triggering unidirectional fluid secretion from exocrine pancreas. Nature, 348,735-738.

9. Parker I. Ivorra I. (1990) Localized all-or-none calcium liberation by inositol trisphosphate. Science, 250,977-979.

10. Hemsndez-Crux A. Sala F. Adams PR (1990) Subcellular calcium transients visualized by confocal microscopy in a voltage-clamped vertebrate neuron. Science, 247.858-862.

11. Pawley JB. (1990) Handbook of Biological Confocal Microscopy. New York, Plenum Press.

12. Wilson T. (1986) Confocal light microscopy. Ann. NY Acad. Sci., 483,416-427.

26.

27.

28.

Kauffman RF. Taylor RW. Pfeiffer DR. (1980) Cation transport and specificity of ionomyckc comparison with ionophore A23187 in rat liver mitochondria. J. Biol. C&m., 255,2735-2739. Yamaguchi Y. Dalle-Molle E. Hard&m WGM. (1991) Vasopressin and A23 187 stimulate phosphorylation of myosin light chain-l in isolated rat hepatocytes. Am. J. Physiol., 261. G312-0319. Watanabe N. Tsukada N. Smith CR. Phillips MJ. (1991) J. Cell. Biol., 113, 1069-1080.

13. Graf J. Boyer JL. (1990) The use of isolated rat hepatccyte couplets in hepatobillary physiology. J. Hepatol., 10, 387-394.

14. Gautam A. Ng OC. Strazaabosco M. Boyer JL. (1989) Qusntitative assessment of canalicular bile formation in isolated hepatocyte couplets using microscopic optical phmimeuy. J. Clin. Invest., 83, 565-573.

Please send reprint requests to : Dr Michael H. Nathanson, Department of Medicine, Room 1080 LMP, Yale University, School of Medicine, New Haven CT 06510, USA.

Received : 19 August 1991 Revised : 3 October 1991 Accepted : 14 October 1991