Paracrine effects of bombesingastrin-releasing peptide and other growth factors on pulmonary neuroendocrine cells in vitro.код для вставкиСкачать
THE ANATOMICAL RECORD 236:53-61 (1993) Paracrine Effects of BombesidGastrin-Releasing Peptide and Other Growth Factors on Pulmonary Neuroendocrine Cells In Vitro VALERIE SPEIRS, ELIZABETH BIENKOWSKI, VERONICA WONG, AND ERNEST CUTZ Department of Pathology and The Research Institute, The Hospital for Sick Children, Toronto, Ontario, Canada ABSTRACT Pulmonary neuroendocrine cells (PNEC) are numerous in the fetus where they have been implicated to have a role in fetal lung development. We assessed the effects of putative growth factors, gastrin releasing peptide (GRP), cholecystokinin (CCK), gastrin (GN), serotonin (5HT), and epidermal growth factor (EGF), some of which are produced by PNEC, either alone or in combination, on cultured fetal rabbit PNEC from 20,24, and 28 day fetuses. GRP increased the total protein of the cultures over a 7 day period in an age-dependent manner, with greatest effect in cultures from the 24 day fetus, no effect with the 28 day fetus, and an inhibitory effect on 20 day cultures. This was accompanied by an increase in PNEC, which could be blocked by treatment of the cultures with a monoclonal antibody to GRP (2All). There was no increase in 3H-thymidine labeling of PNEC in GRP treated cultures but an increase in numbers of cells partially stained for 5-HT, suggesting the induction of a precursor cell. Other growth factors had neither an inhibitory nor a stimulatory effect either alone or in combination with GRP. Preliminary studies with lZ5IGRP receptor localization suggests that the GRP receptor is mostly expressed on pulmonary fibroblasts, and less on epithelial cells, so that the role for GRP in fetal lung development, at least in the rabbit, is probably indirect, acting via a paracrine mechanism. o 1993 Wiley-Liss, Inc. Key words: Neuroendocrine cells, Bombesin, Gastrin-releasing peptide, Cholecystokinin, Gastrin, Serotonin, Autocrine and paracrine growth, Receptors Control of development in fetal lung is governed by growth and differentiation factors which act locally in a paracrine or autocrine fashion. Pulmonary neuroendocrine cells (PNEC), a sparsely distributed cell population within the airway mucosa, are believed to have a role in lung development, because in most species, they are present in highest levels just prior to birth and decline postnatally, such that in the adult, they are present only in extremely low numbers (Cutz et al., 1984; Cho et al., 1989). Many peptides have been implicated to have a role in fetal lung development, the most well studied being gastrin releasing peptide (GRP). GRP, the mammalian homologue of the amphibian neuropeptide bombesin, represents one of a family of small peptides which are biologically active in normal and malignant cells. This peptide has been shown to be a growth promoter for a number of different cell types including 3T3 fibroblast (Rosengurt and SinnettSmith, 1983; Wakelam et al., 1983), bronchial epithelial cells (Willey et al., 1984), human and mouse fetal lung (Spindel et al., 1987; Sunday et al., 19901, and human breast carcinoma cells (Nelson et al., 1991). Perhaps the most well-studied role of GRP is its autocrine growth regulation of human small-cell lung car8 1993 WILEY-LISS, INC. cinoma (SCLC) (Cuttitta et al., 1985; Weber et al., 1985). GRP also stimulates the growth of SCLC cells in a clonogenic assay (Carney et al., 1987) and the growth of SCLC xenografts in nude mice (Alexander et al., 1988; Sehti and Rozengurt, 1991), and in vitro and in vivo growth of SCLC in nude mice is greatly inhibited by a monoclonal antibody ( 2 A l l ) against GRP (Cuttitta et al., 1985). PNEC share certain features with their neoplastic counterpart, SCLC, notably APUD cell properties (amine precursor uptake and decarboxylation; Pearse, 1969). Since GRP has been demonstrated to be a n autocrine growth regulator for SCLC in vitro, we examined the possible autocrine role of bombesin and other regulatory peptides on PNEC in fetal lungs using cultures of these cells. As well as looking a t the effects of GRP, we investigated the role of other potential growth factors, some of Received November 28, 1991; accepted April 28, 1992. Address reprint requests to E. Cutz, MD, Department of Pathology, The Hospital for Sick Children, 555 University Avenue, Toronto M5G 1x8, Ontario, Canada. 54 V. SPEIRS ET AL. which are produced by PNEC. Cholecystokinin (CCK), a calcium mobilizing neuropeptide, can function as a growth factor for human SCLC (Sehti and Rozengurt, 1991) and cultured colon tumor cells (Hoosein et al., 1990) and CCK-like immunoreactivity has been demonstrated in fetal rabbit PNEC (Wang and Cutz, 1990; see also this issue), suggesting it may play a n autocrine role. Gastrin (GN), which has similar properties to CCK, and is one of the peptides released by the action of GRP in the gut (Bertaccini et al., 1974; McDonald et al., 1979; Spindel, 1986) was included as well. Serotonin (5-HT) is a biogenic amine which is produced from the essential amino acid, tryptophan and although not strictly speaking a growth factor, was also included in the study. There are appreciable amounts of 5-HT in the fetalineonatal lung, found mainly in PNEC but also in mast cells and platelets (Keith et al., 1981; Lauweryns et al., 1982). 5-HT has been shown to be mitogenic under certain circumstances and the role of 5-HT as a growth factor has recently been proposed (Seuwen and Pouyssegur, 1990). Epidermal growth factor (EGF), a pleitrophic polypeptide which has been implicated to play a role in normal growth and development, including the lung (Stahlman et al., 19891, was also studied. MATERIALS AND METHODS Cell Isolation The cell isolation procedure was essentially the same a s our previously published method (Cutz et al., 1985) but with the following modifications. New Zealand White does (Reimans Fur Ranch, Guelph, Ontario, Canada) of 20, 24, or 28 days gestation were sacrificed by lethal intravenous injection of sodium pentobarbital. The fetuses were removed and the entire lungs aseptically dissected out, washed three times with serum-free a-minimal essential medium (a= MEM; University of Toronto, Toronto, Ontario, Canada), and chopped into 1-2 mm3 fragments using crossed scalpel blades. For the 20 and 24 day fetuses, the fragments were incubated with 0.1% collagenase type IV (Sigma Chemical Company, St. Louis, MO) for 60 minutes a t 37°C with shaking, washed three times with serumfree a-MEM and any obvious connective tissue removed, then incubated for a further 30 minutes with 0.25% pronase with 0.001% DNase (both from Sigma) to yield a single cell suspension. Lungs from the 28 day fetuses were minced as above and incubated for 30 minutes with 30 ml of prewarmed 0.3% collagenase (Sigma) a t 37°C. The epithelial fragments were pelleted and incubated for a further 30 minutes a t 37°C with fresh warm 0.3% collagenase containing 0.1 mgiml elastase (Sigma). After incubation, the fragments were pipetted until they were small enough to pass easily through a 10 ml pipette. The remainder of the protocol was identical to that described for the 20 and 24 day fetuses. Two-milliter aliquots of single cell suspensions from each fetal age group were layered on top of a 10 ml continuous gradient of 50% Percoll (Pharmacia, Dorval, Quebec, Canada), and centrifuged a t 2,000 rpm for 10 minutes a t 4°C. PNECs concentrated a t the surface of the gradient and were aspirated using a 5 ml pipette. Cell viability, assessed by trypan blue exclusion, always exceeded 95%. Cell Culture Primary cultures of fetal rabbit lung were initially maintained in a-MEM containing the following supplements: 5 p,g/ml bovine insulin, 5 p.g/ml hydrocortisone, 1 p,g/ml retinol acetate (all from Sigma), 50 unitsiml penicillin/streptomycin (both Gibco, Burlington, Ontario, Canada) in the presence of 10% fetal bovine serum (Gibco), in Lab Tek chambers, 4-well plates or 60 mm Petri dishes (all Gibco). After 24 hours the serum concentration was reduced to 0.5% to retard fibroblast overgrowth; for experiments involving growth factors, the serum concentration was 0.5% throughout. Growth Factors The following growth factors, either alone or in combination, were added to the cells either from the initiation of the culture or after a period in culture, a t the following concentrations: bombesidGRP (0-100 ngiml; Sigma #G8022), CCK 0-100 ng/ml; Sigma #C2901, fragment 26-33), GN (Sigma, #G5024, big gastrin 11, and EGF (0-10,000 ngiml; Sigma). 5-HT (0.01 mM-10 mM; Sigma), although not strictly speaking a growth factor, was included in the study. For specific details of the growth factor combinations, refer to individual experiments. Following culture, the cells were either assayed for protein content or immunostained for 5-HT and the numbers of 5-HT positive cells in a given area counted. In selected experiments, we added a n IgGl monoclonal antibody to bombesin ( 2 A l l ; Boehringer Mannheim Biochemicals, Dorval, Quebec, Canada) a t dilutions of 1:1,000 or 12300. 3H-Thymidine Labeling Selected cultures were rinsed three times with a-MEM (Gibco) and pulse labeled for 1 hour a t 37°C with 2.5 PCiiml of 3H-thymidine (Amersham, Oakville, Ontario, Canada) a s previously reported (Cutz et al., 1985). The medium was aspirated and the cells washed three times with phosphate buffered saline (PBS), fixed with 10% neutral buffered formalin (NBF), and immunostained for 5-HT. Slides were then coated with Kodak nuclear track emulsion, exposed for 1 week a t 4"C, and developed according to standard procedures. Receptor Autoradiography This was performed on unfixed cultures according to a modification of O'Donohue et al. (1984). Three day cell cultures were washed with three changes of PBS containing 0.07% bovine serum albumin (BSA) then incubated with 250 Ciimmol (341251]-iodotyroso115)GRP (Amersham), in the presence of 1 pM cold GRP (Sigma) for either 10, 20, 30, 40, 50, or 60 minutes a t room temperature. At the end of the incubation, each slide was washed extensively with PBS, coated with photographic emulsion (Kodak nuclear track), exposed for 48 hours at 4"C, then developed in Kodak D19 for 4 minutes followed by a 5 minute fixation. lmmunohistochemistry For 5-HT immunostaining, cells were fixed in 10% NBF for 1 hour a t room temperature then immunostained with a rat monoclonal antibody to 5-HT (SeraLab, UK) according to Cutz et al. (1985). For cytokeratin immunostaining, the cells were fixed as above, 55 EFFECTS O F GROWTH FACTORS ON PNEC 50 1 T 40 30 Control 1 ng/ml 20 1Ong/rnl 100ng/rnl 10 0 1 2 3 Day of 4 culture 5 Fig. 1. Effect of GRP on total protein content of cultures from 24 day rabbit fetal lung. "_ H Control lng/mi 1Ong/ml 100 0 Day 1 Day 3 Jt- p<o.002 ** p<0.00! Day 7 Days in culture Fig. 2. Effect of GRP on PNEC numbers in 24 day fetal cultures. incubated with CAM 5.2 (Becton Dickinson Inc., Mississauga, Ontario, Canada) for 30 minutes at room temperature followed by horseradish peroxidase (HRPI-conjugated secondary antibody, and visualized by immersing in 0.5 mg/ml diaminobenzidine containing 0.05% hydrogen peroxide. High Performance Liquid Chromatography HPLC was performed on cell extracts of GRP treated cultures and some extracts of cultures treated with other growth factors. Three day cultures growing in 60 mm dishes were washed three times with distilled water then extracted with 1 ml of 0.1 M perchloric acid (BDH Inc., Toronto, Ontario, Canada) containing 2.7 mM ethyldiaminetetraacetic acid (EDTA) (BDH) and 0.2 mM 3,4-dihydroxybenzylamine(Bioanalytical Sys- tems, West Lafayette, IN) a s a n internal standard, in the dark at 4°C overnight. Cell extracts were filtered through 0.22 pm filters (Millipore, Mississauga, Ontario, Canada) and applied to a C18 HPLC column (Waters, Mississauga, Ontario, Canada). The mobile phase consisted of 0.1 M sodium acetate (BDH), 0.12 mM EDTA (BDH), 0.816 mM sodium octyl sulphate (Bioanalytical Systems), and 25% glacial acetic acid (BDH) made up in 20% methanol (HPLC grade; Caledon Laboratories, Georgetown, Ontario, Canada). Flow rate was 1 ml per minute and the pressure was 11.1 kPa. Protein Determination Protein content was estimated using the Bradford protein assay (Bradford, 1976; BioRad). V. SPEIRS ET AL. 56 -r 120 T ?! CI a, .-E CI C a, u Control 60 t n 1Ong/rnl GRP 40 0 W z n 20 0 -r Day 1 Day 3 Day 7 In culture Days Fig. 3. Effect of GRP on PNEC numbers in 28 day fetal cultures. .- e, C a, 20 0 T - Control rn lnglml GRP 10ng/ml GRP L 10 - P) n 0 W z a 0- 1 T Day 3 Day 7 Days in culture Fig. 4. Effect of GRP on PNEC numbers in 20 day fetal cultures. Statistics All statistical analysis was by the Student's t-test, unpaired test. Fa ESULTS Age-Related Effect of GRP on PNEC Cultures Cell cultures from 20, 24, and 28 day fetuses were treated with daily doses of either 1 , 1 0 , or 100 ngiml of GRP and the total protein content measured on a daily basis the Bradford assay. Cultures from the 24 fetus showed a dose-related increase in total protein over the 7 day culture period in the presence of GRP (Fig. 11, with maximum effect observed in the presence of 10 ng/ml of GRP on day 7. No effect on total protein was observed in GRP treated cultures from 20 and 28 day fetuses during the 7 day period (data not shown). EFFECTS OF GROWTH FACTORS ON PNEC 57 TABLE 1. H P L C determination of 5-HT levels in cell extracts of 3 d a y cultures of rabbit fetal lung' Mean 5-HTi25 Fling protein ( t S E ) 70 t 10.37(n=4) 130 t 5.77(n= 3)* Control 10 ng/ml GRP 'Cells were incubated with 10 ng/ml GRP for 3 days then 5-HT content measured as described in the Materials and Methods. *P < 0.002 compared to control, Student's t-test, unpaired samples. TABLE 2. Inhibition of rabbit P N E C increases after incubation with a monoclonal antibody to bombesin' Experiment Control 1 ngiml GRP MAb 10 ngiml GRP + MAb Control 1 ngiml GRP MAb 10 ngiml GRP + MAb + + An ti bodv concentra6on 1:l.OOO 1:1;000 1:1,000 1:800 1:800 12300 Mean PNEC (+-SE) 365.87 t 14.25 392.62 t 39.12 375.25 26.25 297.75 f 18.25 312.37 t 29.62 292.37 t 4.62 * 'Cultures treated with GRP were incubated with 2 A l l monoclonal antibody (MAb) in Lab Tek chambers at the dilutions indicated for 3 days, after which the cells were stained for serotonin and the numbers of positive cells counted. TABLE 3. Effect of short G R P incubations on rabbit PNEC in vitro' Day Day Day Day Dav of treatment 2-day 3 2-day 3 7-day 8 7-dav 8 GRP concentration (neiml) 1.0 10.0 1.0 10.0 % increase 6 6 <170 <l% 'Cells were incubated with GRP after either 2 or 7 days in culture, fixed, stained for 5-HT, and the numbers of positively stained PNEC expressed a s a percentage of untreated controls. To assess the effects of GRP on PNEC number, treated cultures were fixed and immunostained with a monoclonal antibody against 5-HT and the numbers of Fig. 5. A typical preparation showing a partially stained PNEC positive cells in a given area counted. Cultures from (arrow). Note the close location of a fully stained PNEC. Preparations the 24 day fetus showed a significant 30% increase in were imrnunostained for 5-HT. x 9500. PNEC number over control levels by day 3, in the presFig. 6. Autoradiograph of 7 day culture pulse labeled with 3H-thyence of 10 ng/ml GRP; this effect was sustained up to midine for 1 hour, fixed, immunostained for 5-HT, and developed. day 7 (Fig. 2). This effect was abolished with the 28 day Note the heavy concentration of silver grains over the nucleus of a fetus (Fig. 31, while in the GRP containing cultures PNEC positively stained for 5-HT (arrow) x 500. from the 20 day fetus there was some suggestion of a n inhibitory effect (Fig. 4). HPLC determination of extracts from treated 24 day fetuses showed a n approxiPartial Staining Effect mate doubling in 5-HT levels over controls (Table 1). The GRP effect observed in the 24 day fetal cultures Whilst examining the GRP treated cultures, we nocould be effectively blocked by the addition of a mono- ticed that some of the PNEC were only partially clonal antibody against GRP ( 2 A l l ) to the cultures stained with 5-HT (Fig. 5).At first we thought this was (Table 2). artefact, but the partially stained cells were usually situated adjacent to fully stained PNEC. A partially Short Term GRP Incubations stained cell was defined as a cell which showed a granIn some of the 24 day fetal cultures, instead of adding ule-like staining pattern or staining at the cell periphGRP from the initiation of culture as above, the cells ery, whilst a fully stained cell exhibited a dark and were cultured in the absence of GRP for either 24 hours even cytoplasmic immunoreactivity. Counts of partial or 1week, then the growth factor added for a 24 hour and fully stained cells was performed using several period, and the numbers of 5-HT positive cells counted. representative slides and the results are summarized In both cases, the GRP effect was lost (Table 3). in Table 4. In the presence of GRP, the ratio of fully to 58 V. SPEIRS ET AL. TABLE 4. Effect of GRP on partial staining of PNEC from 24 day fetal rabbit lung (3 day cultures) Mean no. 5-HT positive PNEC (2SE)' 169.87 t 9.85 36.68 2 4.75 255.87 ? 24.3 96.18 4.24 * Staining pattern' (full or partial) Full Partial Full Partial * GRP (10 ngiml) Ratio of fully: partially stained PNEC - + + 5:1 3:1 'Number of 5-HT positive cells per square centimeter of a Lab Tek chamber. 'Full staining was determined by a n intensely and uniformly stained cytoplasm while partially stained cells showed weak cytoplasmic staining or staining a t the cell periphery (see Fig. 5). servations of Sunday et al. (19901, who showed that in fetal mouse lung, there was a critical window period 2-4 days towards the end of gestation when the GRP H- Thymidine Incorporation effect could be observed. Related to this, in our experThere was no increase in labeling index of PNEC in iments, GRP had to be present from the start of culture the presence of GRP compared to controls. Labeled in order to elicit a response, a s its effect was lost in 24 PNEC (Fig. 6) were only observed towards days 5-7 of day fetal cultures, cultured in the absence of the culture. The estimated labeling index of PNEC was growth factor for a period, then treated with GRP for zero from days 1-5 of culture and up to 5% from days the final 24 hours of culture. There was a variation in absolute numbers of PNEC 5-7. However, there was a modest and sustained increase in labeling index of the non-neuroendocrine cell from the different gestational ages used. Highest numpopulation (predominantly fibroblasts) throughout the bers of PNEC were observed in the 24 day fetus while culture period in the presence of GRP (data not shown). the numbers from the 20 and 28 day fetuses were 10fold and %fold lower, respectively. This is consistent Effects of Other Growth Factors with previous observations which show a prenatal peak We assessed the effect of a number of other growth followed by a rapid postnatal decline of PNEC in most factors, either alone or in combination with GRP, on species (Cutz et al., 1984; Cho et al., 1989). Although one or more of the following parameters: total protein absolute PNEC number was extremely low in cultures content, PNEC number, and 5-HT content by HPLC. from the 20 day fetus, which resulted in high standard CCK, GN, and EGF had no effect on any of these pa- errors, nevertheless, there was a suggestion of a n inrameters a t the concentrations used. Incubation of cul- hibitory effect in the presence of GRP. conceivably, tures with 5-HT increased, as expected, the amount of there could be mechanisms to prevent these cells from 5-HT detected by HPLC and increased the background responding to GRP a t such a n early age, perhaps in the immunostaining. Combining GRP (10 ng/ml) with dif- form of a n inhibitor. We tried using higher concentraferent concentrations of either EGF (1-10,000 ng/ml) tions of GRP (data not shown) but still failed to elicit a or CCK (1-100 ng/ml) showed neither a greater or response. There was no increase in 3H-thymidine labeling of lesser effect than with GRP alone. These results are PNEC in the GRP treated cultures, although there was summarized in Table 5. a modest and sustained increase in labeling of the nonReceptor Autoradiography neuroendocrine population. This can probably be exUsing a simple autoradiography technique combined plained a s a n increase in mesenchymal cell growth, with cytokeratin staining of parallel cultures, prelim- since GRP has been previously demonstrated to be a inary results suggest that GRP receptors are located potent mitogen for fibroblasts (Rozengurt and Sinnettmostly on rabbit fetal lung fibroblasts. Heaviest label- Smith, 1990; Wakelam et al., 1983). Labeled PNEC ing was observed over areas of cultures, which, when were only observed after 5-7 days in culture, indicatcompared with cytokeratin stained parallel cultures, ing that the cells do divide, but have a long population appeared to correspond to fibroblasts (Fig. 7). Labeling doubling time compared with other epithelial cell types was observed after 10 minutes incubation with the ra- (Cutz et al., 1985). The lack of increased labeling in the dioisotope, and sustained for up to 60 minutes. PNEC population after GRP treatment implies that the peptide is not causing cell division in these cells, DISCUSSION but rather having a n indirect effect, perhaps causing a The present study demonstrates that GRP has a n recruitment of either undifferentiated stem cells or image-related effect on rabbit fetal lung growth in vitro, mature PNEC not yet expressing the full PNEC pheand, more specifically, has a n effect on absolute num- notype. This would account for the higher incidence of bers of PNEC. Greatest effect was observed with the 24 partially stained cells observed with the GRP treated day fetal cultures, with no effect with 28 day fetuses, cultures. and some suggestion of a n inhibitory effect with culIt was originally concluded that PNEC represented a tures from 20 day fetuses. This implies that PNEC are terminally differentiated cell population with no furonly responsive to the effects of GRP for a short time ther capacity to divide (Hernadez-Vasqeuz et al., 1978). towards the end of gestation, and agrees with the ob- However, it has subsequently been demonstrated, and partially stained cells was 3:1, in the absence of GRP this rose to 5:l. 59 EFFECTS OF GROWTH FACTORS ON PNEC TABLE 5. Effects of other growth factors on PNECs in vitro' Growth factor(s) Cholecystokinin Gastrin EGF Serotonin GRP cholecystokinin GRP + EGF + Total Drotein No change No change ND No change As GRP alone As GRP alone 5HT staining No change No change No change High background As GRP alone As GRP alone HPLC No change No change No change 10 fold increase ND ND 'Cells were incubated with the above growth factors and the various parameters compared with untreated controls. ND = not done. For details refer to text. Fig 7 a: Autoradiograph of a 3 day unfixed culture incubated with '251-GRP Note the heavy concentration of the label over what appear to be fibroblasts (arrow) x 100. b: A parallel culture of (a),but irnrnunostained for cytokeratin to show islands of epithelial cells (arrow)surrounded by unstained fibroblasts x 220 60 V. SPEIRS ET AL. confirmed in this study that PNEC are a slowly dividing cell population compared with other pulmonary epithelial cells, with a doubling time of around 7 days (Cutz et al., 1985). Induction of PNEC hyperplasia by treating hamsters with the carcinogen, diethylnitrosamine, followed by exposure to 3H-thymidine, then looking at the percentage of 3H-thymidine labeled cells in lung sections, revealed that the epithelial cells immediately adjacent to the NEB were labeled, and not the actual NEB, suggesting that NEB differentiate from the the airway epithelium (Linnoila, 1982). This proposal has been substantiated by Hoyt et al. (1990) who followed the dynamics of NEB formation in hamsters continuously exposed to 3H-thymidine for the final 4.5 days of gestation, and concluded that PNEC arise from differentiation of the adjoining airway epithelial cells. The appearance of partially stained cells in our preparations would lend further support to this proposal. The other growth factors used in this study, CCK, GN, EGF, and 5-HT were neither stimulatory nor inhibitory to the cultures either alone or in combination with GRP, although it is possible they may affect some other parameter we did not evaluate. It is of interest to note that both 5-HT and CCK are present in rabbit PNEC (Cutz et al., 1985; Wang and Cutz, 1990, also see this issue), while GRP immunoreactivity has not bee demonstrated in this species, although i t is abundant in human lung (Wharton et al., 1978; Cutz, 1982; Sunday et al., 1988). Since the cultures did not respond to either CCK or 5-HT, but showed a positive response in the presence of GRP, this demonstrates that the growth factor does not have to be resident within the cell in order for the cell to respond to it. A similar observation has been made with a SCLC cell line, NCI-N417, which displays GRP-dependent growth (Carney et al., 19871, but expressed no detectable mRNA for the GRP receptor (Corjay e t al., 1991). Therefore, in this cell line and in our rabbit model system, GRP would appear to be fulfilling a paracrine rather than an autocrine role. Although the exact source and site of action of GRP in fetal rabbit lung remain to be determined, it would appear that in this species, GRP is acting in a paracrine rather than a n autocrine manner. How this occurs is presently unknown, but based on what happens during development of other pulmonary systems, i.e., the pulmonary surfactant system, we propose the following hypothesis (Fig. 8). It has been demonstrated by Smith (1984) that induction of pulmonary surfactant in fetal lung is indirect. Glucocorticoid binds to a cell surface receptor present on fibroblasts which causes them to release a low molecular weight peptide called fibroblast-pneumocyte factor (Smith, 1978), which in turn stimulates alveolar type I1 pneumocytes to synthesize surfactant (Post and Smith, 1984; Post e t al., 1984). We speculate that a similar type of mechanism may be functioning here, with GRP acting indirectly through the fibroblasts rather than directly on the PNEC. Although abundant in small cell lung cancer (Moody e t al., 19851, there has been only one report of GRP receptors in normal mammalian lung. Using well characterized lZ5I receptor binding studies, Lach et al. (1991) recently demonstrated the presence of high affinity GRP recep- GRP + + + Fetal rabbit lung fibroblasts FACTOR(S) Recruitment of undifferentiated pulmonary stem cells Gradual expression of DCV and 5 H T (partially stained cells) + MATURE PULMONARY NEUROENDOCRINE CELL Fig. 8. Proposed sequence of events mediated by GRP in rabbit lung. tors in guinea pig lung membranes, so it is likely that they are present in other species a s well. Although we have yet to demonstrate GRP receptors on rabbit lung membranes, our preliminary data with 1251-GRPautoradiography on whole cells, accompanied by cytokeratin staining of parallel cultures, suggest that GRP receptors are located mostly on fibroblasts, and less on epithelial cells. Moreover, the fact that we were able to block the GRP effect using a monoclonal antibody to GRP supports the role of a GRP receptor in rabbit fetal lung. The binding of GRP could bring about the release of another, as yet unidentified factor, produced by fibroblasts. This may act on undifferentiated stem cells, causing them to gradually acquire PNEC features, e.g., synthesis of 5-HT; this would account for the partially stained PNEC population. The paracrine effects of GRP on normal rabbit PNEC stands in contrast to the autocrine role observed with the majority of SCLC cell lines in vitro. This suggests that different regulatory mechanisms operate during normal lung development compared to pulmonary neoplastic transformation process. The acquisition of GRP receptors resulting in activation of autocrine stimulation loop may be a particular feature associated with neoplastic transformation of pulmonary neuroendocrine-derived cells. A note of caution is in order since the present study concentrated only on possible growth promoters and none of the growth suppressing agents or differentiating factors which are known to be present in vivo (Deuel, 1987). Future studies will concentrate on further defining the role of peptides in this system a t the cellular and molecular level, a s well as further identifying GRP binding sites in rabbit fetal lung and other species. EFFECTS OF GROWTH FACTORS ON PNEC ACKNOWLEDGMENTS This work was supported by grants from the Medical Research Council of Canada (PG-42) and NICHD (1 R01 HD22713). V.S. is a recipient of a n MRC postdoctoral fellowship. LITERATURE CITED Alexander, R.W., J.R. Upp, Jr., G.J. Poston, V. Gupta, C.M. Townsend, Jr., and J.C. Thompson 1988 Effects of bombesin on growth of human small cell lung carcinoma in vivo. Cancer Res., 48t1439-1441. Bertaccini, G., V. Erspamer, P. Melchiorri, and N. Sopranzi 1974 Gastrin release by bombesin in the dog. Br. J. Pharmacol., 52t219225. Bradford, M. 1976 A rapid and sensitive method for the quantitation of microgram quantities of protein using the principles of protein dye binding. Anal. Biochem., 72:248-254. Carney, D.N., F. Cuttitta, T.W. Moody, and J.D. Minna 1987 Selective stimulation of small cell lung cancer clonal growth by bombesin and gastrin-releasing peptide. Cancer Res., 47:821-825. Cho, T., W. Chan, and E. Cutz 1989 Distribution and frequency of neuroepithelial bodoes in post-natal rabbit lung: Quantitative study with monoclonal antibody against serotonin. Cell Tissue Res., 255:353-362. Corjay, M.H., D.J. Dobrzanski, J.M. May, J . Viallet, H. Shapira, P. Worland, E.A. Sausville and J.F. Battey 1991 Two distinct bombesin receptor subtypes are expressed and functional in human lung carcinoma cells. J Biol Chem, 266: 18771-18779. Cuttitta, F., D.M. Carney, J . Mulshine, T.W. Moody, J . Fedorko, A. Fischer, and J.D. Minna 1985 Bombesin-like peptides can function as autocrine growth factors in human small-cell lung carcinoma. Nature, 316r823-826. Cutz, E. 1982 Neuroendocrine cells of the lung: An overview of morphologic characteristics and development. Exp. Lung Res., 3t185-208. Cutz, E., J.E. Gillan, and N.S. Track 1984 Pulmonary endocrine cells in developing human lung and during neonatal adaptation. In: The Endocrine Lung in Health and Disease. K.L. Becker, A.F. Gazdar, eds. W.B. Saunders, Philadelphia, pp. 210-221. Cutz, E., H. Yeger, V. Wong, E. Bienkowski, and W. Chan 1985 In vitro characteristics of pulmonary neuroendocrine cells from rabbit fetal lung. I Effects of culture media and nerve growth factor. Lab. Invest., 53t672-683. Deuel, T.F. 1987 Polypeptide growth factors: Roles in normal and abnormal cell growth. Annu. Rev. Cell Biol., 3:443-492. Hernadez-Vasquez, A,, J.A. Will, and W.B. Quay 1978 Quantitative characteristics of the Feyter cells and neuroepithelial bodies of the fetal rabbit lung in normoxia and chronic short term hypoxia. Cell Tissue Res., 189r179-186. Hoosein, N.M., P.A. Kiener, R.C. Curry, and M.G. Brattain 1990 Evidence for autocrine stimulation of cultured colon tumor cells by a gastrinicholecystokinin-like peptide. Exp. Cell Res., 186:15-21. Hoyt, R.F., Jr., N.A. McNelly, and S.P. Sorokin 1990 Dynamics of neuroepithelial body (NEB) formation in developing hamster lung: Light microscopic autoradiography after 'H-thymidine labeling in vivo. Anat. Rec., 227t340-350. Keith, I.M., L. Wiley, and J.A. Will 1981 Pulmonary neuroendocrine cells: Decreased serotonin fluorescence and stable argyrophil-cell numbers in acute hypoxia. Cell Tissue Res., 214t201-207. Lack, E., A. Trifilieff, Y. Landry, and J.-P. Gies 1991 High-affinity receptors for bombesin-like peptides in normal guinea pig lung membranes. Life Sci., 48:2571-2578. Lauweryns, J.M., V. de Bock, A.A.J. Verhofstad, and H.W.M. Steinbusch 1982 Immunohistochemical localization of serotonin in intrapulmonary neuroepithelial bodies. Cell Tissue Res., 226215219. Linnoila, R.I. 1982 Effects of diethylnitrosamine on lung neuroendocrine cells. Exp. Lung Res., 3r225-236. McDonald, T.J., H. Jornvall, G. Nilsson, M. Vagne, M. Ghatei, S.R. Bloom, and V. Mutt 1979 Characterization of a gastrin-releasing peptide from porcine non-antral gastric tissue. Biochem. Biophys. Res. Commun., 9Or227-231. 61 Moody, T.W., D.N. Carney, F. Cuttitta, K. Quattrochi, and J.D. Minna 1985 High affinity receptors for bombesidGRP-like peptides on human small cell lung cancer. Life Sci., 37t105-113. Nelson, J., M. Donnelly, B. Walker, J . Gray, C. Shaw, and R.F. Murphy 1991 Bombesin stimulates proliferation of human breast cancer cells in culture. Br. J. Cancer, 63:933-936. ODonohue, T.L., V.J. Massari, C.J. Pazoles, B.M. Chronwall, C.W. Schults, R. Quiron, T.N. Chase, and T.W. Moody 1984 A role for bombesin in sensory processing in the spinal cord. J. Neurosci., 12t2956-2962. Pearse, A.G.E. 1969 The cytochemistry and ultrastructure of polypeptide hormone-producing cells of the APUD series and the embryologic, physiologic and pathologic implications of the concept. J. Histochem. Cytochem., 17.303-313. Post, M., and B.T. Smith 1984 Effect of fibroblast-pneumocyte factor on the synthesis of surfactant phospholipids in type I1 cells from fetal rat lung. Biochim. Biophys. Acta, 793t297-299. Post, M., J.T. Torday, and B.T. Smith 1984 Alveolar type I1 cells isolated from fetal rat lung organotypic cultures synthesize and secrete surfactant-associated phospholipids and respond to fibroblast pneumonocyte factor. Exp. Lung Res., 7r53-65. Rozengurt, E., and J. Sinnett-Smith 1990 Bombesin stimulation of fibroblast mitogenesis: Specific receptors, signal transduction and early events. Phil. Trans. R. SOC.Lond., B327:209-221. Sehti, T., and E. Rozengurt 1991 Multiple neuropeptides stimulate clonal growth of small cell lung cancer: Effects of bradykinin, vasopressin, cholecystokinin, galanin and neurotensin. Cancer Res., 51:3621-3623. Seuwen, K., and J . Pouyssegur 1990 Serotonin as a growth factor. Biochem. Pharmacol., 39r985-990. Smith, B.T. 1978 Fibroblast-pneumocyte factor: Intercellular mediator of glucocorticoid effect on fetal lung. In: Neonatal Intensive Care. I Stern, H. Ow, and B. Friis-Hansen, eds. Masson, New York, Vol. 11, pp. 25-32. Smith, B.T. 1984 Pulmonary surfactant during fetal development and neonatal adaptation: Hormonal control. In: Pulmonary Surfactant. B. Robertson, L.M.G. Van Golde, and J.J. Batenburg, eds. Elsevier, Amsterdam, pp. 357-381. Spindel, E.R. 1986 Mammalian bombesin-like peptides. Trends Neurosci., 9:130-139. Spindel, E.R., M.E. Sunday, H. Hofler, H.J. Wolfe, J.F. Habener, and W.W. Chin 1987 Transient elevation of messenger RNA encoding gastrin-releasing peptide, a putative pulmonary growth factor in human fetal lung. J. Clin. Invest., 80:1172-1179. Stahlman, M.T., D.N. Orth, and M.E. Gray 1989 Immunocytochemical localization of epidermal growth factor in the developing human respiratory system and in acute and chronic lung disease in the neonate. Lab. Invest., 60,539-547. Sunday, M.E., J. Hua, H.B. Dai, A. Nusrat, and J.S. Torday 1990 Bombesin increases fetal lung growth and maturation in utero and in organ culture. Am. J. Respir. Cell Mol. Biol., 3:199-205. Sunday, M.E., L.M. Kaplan, E. Motoyama, W.W. Chin, andE.R. Spindel 1988 Biology of disease: Gastrin-releasing peptide (mammalian bombesin) gene expression in health and disease. Lab. Invest., 595-24. Wakelam, M.J.O., S.A. Davies, M.D. Houslay, I. Mackay, C.J. Marshall, and A. Hall 1983 Normal p21N-ras couples bombesin and other growth factor receptors to inositol phosphate production. Nature, 323r173-176. Wang, Y.-Y., and E. Cutz 1990 Cholecystokinin (CCK)-like immunoreactivity in neuroendocrine (NE) cells of mammalian lungslight and EM immunohistochemical study. Lab. Invest., 62t109 (abstract). Weber, S., J.E. Zuckerman, D.G. Bostwick, K.G. Bensch, B.I. Sikic, and T.A. Raffin 1985 Gastrin releasing peptide is a selective mitogen for small cell lung carcinoma in vitro. J. Clin. Invest., 75: 306-309. Wharton, J., J.M. Polak, S.R. Bloom, M.A. Ghatei, E. Solcia, M.R. Brown, and A.G.E. Pearse 1978 Bombesin-like immunoreactivity in the lung. Nature, 273:769-770. Willey, J.C., J.F. Lechner, and C.C. Harris 1984 Bombesin and the C-terminal tetradecapeptide of gastrin-releasing peptide are growth factors for norma 1human bronchial epithelial cells. Exp. Cell Res., 153:245-248. 62 WORKSHOP ON PULMONARY NEUROENDOCRINE CELLS DR. POLAK: Questions for both speakers. A very nice presentation. These effects of bombesin on 5-HTcontaining cells, that’s one matter, but if bombesin or its message is not there, that’s another. DR. SPEIRS: Well, we used Northern blots to look for GRP mRNA in fetal lung tissue and cultures, but we haven’t found the message. It could well be that bombesin isn’t expressed. Perhaps it’s a molecule with similar actions but sufficiently different that it isn’t detected by antibodies to bombesin. DR. POLAK: Maybe. And is this rabbit tissue? DR. SPEIRS: Right. DR. POLAK: I’m puzzled by the staining you’ve done, because in our experience in staining rabbit tissue we don’t use rabbit antibodies. It’s impossible. DR. CUTZ: Now, the CCK that I showed, that’s with the rabbit anti-CCK. That’s why there is background. DR. POLAK: So you are one of the very few that can stain rabbit tissue with a rabbit antibody! Most of us can’t. DR. SCHEUERMANN: I fully agree, Dr. Polak. But the same experiments have been effectuated by Dr. Lauweryns also with anti-serotonin antibodies raised in rabbits against rabbits. I cannot understand. DR. CUTZ: But those do stain, in fact. DR. SCHEUERMANN: But I agree fully with Dr. Polak. DR. SUNDAY: I just have one comment though. In the mouse fetal lung we never really saw GRP message on total RNA. There was a only a vague hint of a band. DR. SPEIRS: Yes. Well, this was total RNA that we used. DR. SUNDAY: We had to go to polyA+ RNA to see anything. DR. SPEIRS: Is that right? DR. SUNDAY: It’s different. In human fetal lung you can see a very good band on total RNA, but it’s a question of the relative abundance. Perhaps you could use reverse transcribed RNA PCR detection, that’s another way, or S1-nuclease, except that we don’t have a homologous probe for the rabbit. DR. LINNOILA: Or in situ. DR. SUNDAY: Oh, in situ also. Yes. DR. SCHEUERMANN: Firstly, to Dr. Speirs. In the last experiments with glucocorticoid administration, you state that there is a n indirect influence on pneumonocyte-2 through fibroblasts, but I cannot agree. In my experiments many years ago I demonstrated a direct influence of glucocorticoids on pneumonocyte-2. After administration of this drug there is a n increase of rough endoplasmic reticulum in these cells accompanied by a decrease of glycogen by 25 percent, and exocytosis of the surface lining material. There was also a n increase of lung neuroepithelial bodies. I have another question for Dr. Cutz. You demonstrate very well after hypoxia the release of bioactive substances by uninnervated cells, and my question to you is what is the influence of the nerve endings on exocytosis by these neuroendocrine cells? Have you any supposition, any hypothesis, any experiments done? DR. CUTZ: Well, i t appears that these cells do not need to be innervated to respond to hypoxia. The same experiment can be done in carotid body. DR. SCHEUERMANN: But as to the neuroepithelial body, is there a modulation by these nerve endings on exocytosis of the neuroendocrine cells? DR. CUTZ: I think there is from Lauweryns’ experiments. I mean he has shown quite clearly that there is modulation. One cannot answer your question using an in vitro system with denervated cells, but, I think, a t the cellular and membrane level, this is the only way you can dissect mechanisms of the stimulusisecretion response. DR. SCHEUERMANN: Yes. But according to the experiments of Dr. Lauweryns, it seems that exocytosis of bioactive substances can only go on by innervation, by modulation of these nerve endings. DR. CUTZ: Well, again, you are dealing with a different situation; in vitro a hypoxic response is clearly obtainable, because we’ve shown that many times. DR. SCHEUERMANN: That’s very interesting. DR. GAIL: Dr. Cutz, can you block that response in vitro with inhibitors or anything? Have you tried? DR. CUTZ: You see, again, what we have to do. Because there is no precedent for these kind of studies on pulmonary neuroendocrine cells, we have to use the carotid body literature as a guide, and I started to go to carotid body meetings to learn. But there are a whole range of things which can be done. You know, they use cyanide, or specific channel blockers, and so on. DR. AGUAYO: I have two questions. Do you know if your anti-bombesin antibody cross-reacts with neuromedin-B or phyllolitorin, and did you look a t bombesin by HPLC and see if the profile was more like GRP or more like other bombesin-like peptides? DR. SPEIRS: We haven’t looked into either of these. DR. STEPHENS: Just to follow up the point that Dr. Cutz was talking about, is it possible to add neural elements to your culture or conditioned medium from nerve cultures to your NEB cultures and see then what is the effect of neural factors on NEB? DR. CUTZ: I guess one could imagine lots of different experiments. No. We haven’t done it. And I don’t know what the answer would be. But certainly these things can be done. DR. STEPHENS: It would be a way of getting a t Dr. Scheuermann’s problem. DR. CUTZ: Yes. DR. BECKER: Just in a generic sense, Dr. Cutz, you talked about the carotid body and dopamine. How about dopamine in the neuroendocrine cell? Has anybody looked a t dopamine? DR. CUTZ: We did look with HPLC for metabolites, and there is no dopamine. There is only serotonin in humans and rats. On the other hand, the carotid body also has serotonin in significant amounts. In fact, in the fetus and newborn there is more serotonin than dopa. DR. BECKER: How about other species as far as dopamine is concerned? DR. SCHEUERMANN: And you say dopamine is never demonstrated in neuroendocrine cells? Well, I have found-but it’s not published-dopaminergic fibers near the neuroendocrine cells. DR. SOROKIN: In which species? DR. SCHEUERMANN: In Polypterus. DR. SUNDAY: What animal? What was it? DR. SCHEUERMANN: Polypterus retropinnis. It’s a bichir. These are descended from palaeoniscid fishes WORKSHOP ON PULMONARY NEUROENDOCRINE CELLS and have paired lungs. I think that these dopaminergic fibers are coming from the microganglia located in the pulmonary interstitial connective tissue. DR. CUTZ: Well, we cannot detect dopamine in the neuroendocrine cells, and it’s very easy to detect it in the carotid body. DR. SUNDAY: We’ve actually been unable to stain for either dopamine or tyrosine hydroxylase in hamster, mouse, or human fetal lung. It doesn’t look as though it’s expressed. But I think your model with hypoxia is very interesting, and I was wondering if you have looked a t thymidine incorporation in the face of hypoxia in your rabbits? DR. CUTZ: Ours are very short-term experiments because you would expect the hypoxic response to happen in milliseconds, rather than larger intervals. We haven’t done experiments where you look at effects of long-term hypoxia in culture, but that would be interesting. DR. MILLER: Have various small cell carcinoma lines been looked a t for hypoxic responses? DR. CUTZ: Yes. Since this is a conference on normal function, we’re not supposed to talk much about this, but we’re doing some studies using these cell lines because they are very easy to patch. Charlotte Newman, who is doing this work, has looked at the H69 line, and these definitely have the features of exocytosis and the ionic channels of excitable cells. About ten percent of the H69 cells show a response to hypoxia, and we’re looking a t some other lines to see what it depends on. As you know, H69 is a very poorly differentiated and poorly granulated cell that is not ideally suited for studies with a neurosecretagogue because the exocytotic machinery is virtually nonfunctional. DR. SCHEUERMANN: I have a n addition to the question about presence of catecholamines in neuroepithelial bodies; yes, I have demonstrated small amounts of catecholamines a s well as serotonin in neuroendocrine cells of neonatal rabbits after administration of the respective amine precursors (Scheuermann et al., 1988). It could be hypothesized, moreover, that animals on a tryptophan-free diet may synthesize phenylethylamines instead of indolethylamines because the aromatic L-amino acid decarboxylase can use either 5-HTP or L-DOPA as a substrate and is therefore, nonspecific. DR. SUNDAY: Just in response to York’s question, a t one point we tried culturing H69, H128 and HZO9sthe latter two have high levels of endogenous bombesin-in the presence of 1 percent oxygen, using a culture system designed for the study of erythropoietin gene expression, and we never saw any change in GRP messenger RNA, for whatever that’s worth. That’s all we have looked at. But i t might be interesting to pursue that by looking a t more parameters. DR. STEPHENS: On a different tack, this is to do with the potassium channels. These are usually calcium-dependent, so they might interact with the raised intracellular calcium levels you saw quite apart from calcium’s other effect. These are also ATP-dependent, presumably because of phosphorylation of the channel itself. Do you have any information about what ATP was doing? DR. CUTZ: As you know, there is a huge literature on it. Some work is going on with the effects of ATP on 63 the carotid body cell, and administered ATP itself, you know, raises the calcium levels in the cells so that there may be some receptors. DR. CUTTITTA: Two questions, Dr. Speirs. Is the cell culture system that you’re growing your cells in, a serum-free system? DR. SPEIRS: It’s 0.5 percent serum. DR. CUTTITTA: 0.5 percent? DR. SPEIRS: Yes. We grow them for the first 24 hours in 10 percent and thenDR. CUTTITTA: But you looked a t a lot of different peptides, or growth factors. Have you ever looked at insulin-like growth factor I or II? DR. SPEIRS: No, we haven’t. DR. CUTTITTA: And in relationship to hypoxia and serotonin, have you ever looked at the enzymes involved with serotonin production, and do the messages go up, or does enzymatic activity actually go -~ up with hypoxia? DR. CUTZ: Again this is something we would be interested in doing but haven’t got to yet. It’s probably easier to do at the molecular level as opposed to just measuring the enzyme. DR. POLAK: But I thought you showed one of the enzymes. Your data showedDR. CUTZ: That’s right. But we don’t have results yet. We have the probe for the tyrosine hydroxylase. DR. JOHNSON: I was just wondering, in terms of the hypoxia experiments, how did you choose your level of oxygen? What I’m getting a t is, do you think there might be a threshold, or is this a graded response? DR. CUTZ: Yes, it is graded-I actually cut a slide from my presentation to show that the response is much less with a PO2 of 50 torr and above, and so it seems to be dependent on the ambient oxygen level. DR. JOHNSON: And did you look a t hyperoxia a t all? DR. CUTZ: We did look, and there was no change with hyperoxia. DR. HOYT: I’d like to return to the central issue of the validity of looking at the effects of hypoxia on the cells in isolation a s opposed to their being innervated. Dr. Lauweryns is unfortunately not here to guide us, but he proposed a scheme on the basis of many, many experiments involving infranodosal vagotomy, supranodosal vagotomy, and survivals for different periods of time, to allow the endings within the neuroepithelial bodies in the lung to degenerate. These led him to believe that the primary response to hypoxia is one at the level of the endocrine cell itself. According to Lauweryns, the initial response is followed by a stimulussecretion coupling that promotes release of the densecored granules mostly across the basal membrane of the cell into the peribronchial connective tissue. But at the same time, all that’s known about the highly developed innervation in these bodies-in the rabbit, mind you, because there is species variation-indicates that some of the granules in the endocrine cells are released from synaptic complexes made with the intraepithelial nerve endings. Lauweryns had the idea that this would cause the nerve ending to depolarize, sending a signal towards the central nervous system. He went on to propose a system of axon collaterals. As the main intracorpuscular afferent ending is stimulated, a wave of depolariza- 64 WORKSHOP ON PULMONARY NEUROENDOCRINE CELLS tion spreads along i t to reach collateral, “efferent-like’’ endings in the base of the epithelium; those endings synapse against the neuroendocrine cell itself and this, Lauweryns believed, is responsible for potentiating further release of granules from the cell base. If you look a t these cells as essentially able to respond independent of their innervation, a t least to a n hypoxic stimulus, then clearly the electrophysiological mechanism is now open to probing and unraveling with Dr. Cutz’s system. Once this is worked out in a refined way, the problem will be to take the information back into the whole lung with the innervation intact and work out the normal neurophysiology. But I think Dr. Cutz has the best way to start, because the primary response, a s best anyone knows, is a t the level of the endocrine cell and the innervation is secondary. I’m also intrigued by those neurite-like outgrowths that develop once the endocrine cells have been put in culture because similar processes are seen in vivo in some species, once these cells begin to differentiate in the epithelium. They extend through the basal lamina, down into the underlying connective tissue compartment. And I know Sergei has taken electron micrographs showing contact being made there with neural elements. It looks as though this may be the preferential site for the formation of associations with motor nerve endings, as opposed to the sensory nerve terminals located higher up, within the epithelium itself. So, the whole idea of getting these cells out reliably where you can actually watch them and patch clamp them, is fascinating. I think it’s a very significant development. DR. SCHEUERMANN: But the definition of neuroepithelial body is a partly innervated structure. DR. HOYT: Yes, except that we all use it very loosely. And the only way to tell whether a body is innervated is to look, and to look at all of it, not just a casual section. The innervation obviously must be significant. DR. McDOWELL: Why is it, by definition, innervated? Because it has “neuro” in its title? DR. CUTZ: As defined by Lauweryns. DR. McDOWELL: But he said that he couldn’t see any innervation in the fetal lung, but the NEBS are there. Did he not say that? DR. SCHEUERMANN: It’s correct what you state, but I thinkDR. McDOWELL: But I mean then I think the definition has to change, doesn’t it? DR. SCHEUERMANN: Maybe change. Yes. DR. HOYT: Well, the neuroepithelial body seems to be the end result of a developmental process that may extend beyond birth in some animals, and the question of a change in the function between the role during development and one played in adult life may hinge, in fact, on the appearance of innervation. DR. SCHEUERMANN: There are so many nerve endings and contacts in these neuroendocrine cells, afferent, efferent, and catecholamine-containing cells, acetylcholinesterase containingDR. HOYT: Like a little peripheral central nervous system. DR. SCHEUERMANN:-immunoreactivity for substance P. There must be modulation of this neuroendocrine cell, that is clear. DR. CUTZ: I think there is repopulation. You know, one is the classical neuroepithelial body of Lauweryns, and the best animal model for that is the rabbit. There is also a second population which, whether called “neuroepithelial body” or “cluster” or whatever, are conglomerates of cells that do not necessarily form a well defined body. And then the single cells. DR. SCHEUERMANN: But the work of Dr. Sorokin is very clear. You have written that neuroepithelial bodies are structures of clear cells innervated, or that have become innervated. That is your definition. DR. CUTZ: Yes. I think that’s true. And when you look a t these bodies, including human, they are indeed innervated. DR. GOSNEY: I think this discussion raises a very important point actually with regards to terminology, because there is, indeed, a n innervated structure which we call a “neuroepithelial body,” which has, I’m sure, as one of its functions, reception. These bodies are quite different from a lot of the clusters of endocrine cells, occurring in diseased lungs in particular, which are not in any way innervated, organized clusters of endocrine cells. I think a lot of confusion creeps in because we all use these terms very loosely. A cluster of endocrine cells is not, as Professor Scheuermann says, a neuroepithelial body, if we use the terminology strictly. DR. McDOWELL: But it might be a n immature one, might i t not? DR. GOSNEY: Well, it might. It’s very difficult to know, if we consider the defining feature is innervation. How many of us ever look for that? We talk about clusters of cells. DR. McDOWELL: Yes. And if their genesis precedes the innervation, the whole thing seems rather academic anyway. Is it a fact that their genesis precedes the innervation? DR. SOROKIN: Yes. If one examines developing lungs in vivo, there first appear one or two uninnervated clear cells, then a cluster of them, and still later an immature, innervated body. If one places a lateembryonic stage lung in organ culture before any clear cells are there, or when just a few are present, structures develop that closely resemble neuroepithelial bodies except for lacking innervation, whatever one may wish to call them. Ingrowth of sensory nerve processes in hilar parts of the lung naturally ceases following surgical interruption of the nerve tracts a t explanation, and intrinsic pulmonary neurons such as postganglionic parasympathetic efferents are almost never found contacting the bodies (for one example, however, see paper by Sorokin et al., Part 2 of these proceedings). In work from Dr. Cutz’s group published in the late 1970s, presumptive sensory nerves innervating neuroepithelial bodies of fetal rabbit lung degenerated soon after organ culturing. DR. HOYT: But the bodies were left behind. DR. SOROKIN: Yes, they were left behind. In sum, the epithelial elements develop first, in vivo as in vitro, and then the nerves grow in. So, if hyperplasia of these cells results from some disease process, for instance, it would seem perfectly logical to find clusters of cells initially. And if those structures survive long enough, who knows, they might become innervated. DR. HUNG: I think in Lauweryns’ initial proposal for the response of neuroepithelial bodies to hypoxia he showed that a t least some of the endocrine cells, or wunnanur UIU ruuviuiuxfil neuroendocrine cells, need to be exposed directly to the air going through the lumen. And as far as I know by EM observations not all the cells in the neuroepithelial bodies are exposed to the lumen. In Dr. Cutz’s experiment using isolated cells, every cell is exposed to hypoxia. So I think we’re probably talking about at least two different populations of cells in this case, some of them exposed to the lumen, which are more receptor type of cells, and other cells underneath them having different functions. DR. SCHEUERMANN: Not all the neuroendocrine cells reach the luminal surface while they are accompanied by Clara-like modified cells which overlap them like roofing tiles. But that is your question. DR. HOYT: But that, again, is a species variable. DR. SCHEUERMANN: Yes. DR. HOYT: I think the experiments of Lauweryns referred to-I wish he were here instead of listening to oral exams, which is what he’s doing right now-were the cross-circulation studies, where he showed that degranulation of the endocrine cells occurred in response to airway hypoxia, not pulmonary arterial hypoxia: That doesn’t really mean that all the cells, to respond to hypoxia, actually have to have a luminal exposure, because it seems logical for the epithelium to receive its oxygen supply from the lumen. After all, in a layer that’s a few micrometers tall a profound hypoxia must make the entire epithelium substantially hypoxic. I know nothing about the hypoxia receptor or whether it’s a cell surface phenomenon. If it’s not, when you lower the oxygen tension in the inspired air you’re actually lowering the oxygen tension in the bronchial lining as a whole, and that means you don’t actually have to have exposure to the bronchial lumen to have this response, whereas with cells in the gut, which respond to chemical stimuli present in the lumen, you really might require a direct exposure to the surface. So, whether or not the lung cells are all open to the lumen or whether they’re all closed, I think, is perhaps not so important. DR. CUTZ: In the fetal neonate rabbit, scanning microscopy has shown that there are large areas exposed to the lumen. I think it’s maybe a misnomer to talk about low blood oxygen level versus air oxygen, because these cells are not in dry air. There is a layer of fluid over them. But the actual oxygen sensor is not known a t the present time. It’s proposed to be a hemelike protein which may be a membrane component and may be cell-specific. DR. HOYT: In human lungs, undoubtedly there are a number of cells which don’t appear to reach the lumen, even allowing for tangential sectioning. I find i t interesting that you see them most often in fetal lungs and in lungs which have been damaged in some way. I just wonder whether some of these are maturing cells which later will reach the lumen. I wonder how many mature cells do actually fail to reach the luminal surface in man. DR. SCHEUERMANN: But it may be possible that some cells, you know, have hypoxia and transmit to the other cells. That is another possibility. DR. SOROKIN: Well, in a study from our lab, Dr. Allan Pearsall reconstructed a neuroepithelial body, as we thought, from serial electron micrographs of adult hamster lung (Pearsall et al., 1985).It turned out to be uninnervated, so probably was still in formation. However, the point is that in single micrographs most of the cells appeared buried and not open to the surface; but from a three-dimensional scale model, one could see that all cells contacted the lumen, if only by a tiny process (see color plate, figures 1 and 2, i n “Lung endocrine cell markers, peptides, and amines,” p . 170.) DR. GOSNEY: I think that’s probably the rule. DR. SOROKIN: I wonder if we can devote the remaining time to more general aspects raised during the session. To start off, I think the thrust of this morning’s presentations is that neuroepithelial bodies or, anyway, pulmonary neuroendocrine cells, have a distinctive prenatal or developmental function. It involves a direct or indirect effect on the proliferation of the endoderm, which apparently is first seen after terminal budding has laid down the airway and it then becomes time for individual airway segments to begin lengthening, in a n orderly accommodation of lung growth to expansion of the thorax. This aspect of neuroendocrine cell function may well predominate until that phase of lung growth nears completion; hence the start-up time and duration of activity could well differ among species according to the length of gestation and the newborn’s relative maturity at birth. In hamsters, activity centers around the time of birth. In the human, I suspect it begins much earlier, perhaps in the second trimester of gestation. I’m not sure. A long-standing question is whether pulmonary neuroendocrine cells decline for a while, if not totally disappear after birth. At least they seem to, where estimates of population size have been made with use of certain markers. Yet we wouldn’t be studying them in adults if they didn’t persist in some form. Perhaps a new population gradually supplants the original, fetal population, some time after birth. It may be appropriate therefore to have a little discussion related to three of the circulated questions listed under “Population Dynamics” (see Issues and Questions Regarding Pulmonary Neuroendocrine Cells, p . 9). Dr. Cutz was interested in the presumed postnatal involution of NEBS, but Dr. Springall thought the question, “Zs there a postnatal decrease in the population size?” should be discussed first. And Dr. Miller had related questions on the same topic. Would anyone care to say something about this? DR. POLAK: Well, regarding whether the cells divide, that has been discussed today quite extensively. So at least we solved that problem. (Laughter.1 DR. POLAK: Changes in phenotypic expression is a big subject on its own because, of course, environmental factors will induce different phenotypes and denervation can induce different phenotypes a s well. So disease states in man or animals will give us a lot of answers on that. DR. SUNDAY: I think the concept of a latent neuroendocrine cell in adults is supported very much by Dr. Polak’s and Springall’s work. There might be a cell that expresses one marker, such as PGP 9.5, and with hypoxic stimulation actually starts to express CGRP. We don’t understand from what cell these arise and what the differentiation sequence is. It would be nice if we could get markers like those for the lymphocyte differentiation pathway to say that the sequence is 66 WORKSHOP ON PULMONARY NEUROENDOCRINE CELLS something like “serotonin occurs very early, bombesin gene expression occurs very early, but then amidation enzymes come in later, as well as calcitonin,” and it seems that information like that will probably come out of the work of people present a t our Workshop. DR. GOSNEY: As to the alleged involution of endocrine cells after birth, personally I feel it’s been rather exaggerated. I think if you go back and look a t the literature very carefully, many of the studies don’t actually take account of the fact that the lungs are growing and that part of this apparent disappearance is merely a dilution amongst the growing airways. I’m sure there is some decrease because they are probably past the one time of life when they are most important, but I think they continue to reside in the lungs right through life. We’ve certainly found them with no problem in people of 70 and 80 years old-very consistently in a very regular pattern (seepaper by Gosney,Part 2 of these proceedings). DR. SPRINGALL: I think I agree with you there, John, but i t also depends on how you actually determine what is and what isn’t a n endocrine cell. DR. GOSNEY: True. DR. SPRINGALL: If you look a t a n endocrine cell by its functional products you can be leading yourself into all sorts of misconceptions because if the cell turns off producing that product, i t still remains a n endocrine cell. It has the ability to synthesize that product. And if there is none there to recognize by immunocytochemical techniques, or whatever, then you’re fooling yourself if you think the number of cells have gone down. It’s simply that they’ve switched off producing that peptide or amine. DR. HOYT: That’s the difficulty in choosing one marker, starting a t day zero and then going through like this. Elizabeth, in fact, showed with the serotonin in the hamsterDR. McD0WELL:-Dr. Keith has shown that too. DR. HOYT: Yes. But you have just shown us pictures-that’s all I meant-where it was elevated on one day, down the next day, and then back again on the following day. And there are others here who have demonstrated a similar pattern. So, you may be misled if you use just one marker. DR. McDOWELL: And furthermore, in the hamster, serotonin was not staining a large number of the bodies, whereas CGRP was. Could I ask you a question, Dr. Cutz? You based your maps of fetal and postnatal rabbit lungs on serial sections. Were they stained with serotonin? DR. CUTZ: There were two markers actually, serotonin and a silver stain. DR. McDOWELL: So the silver stain, do you think that might be picking upDR. CUTZ:-I think that picks up the granules, the chromogranin. DR. McDOWELL: But it looked from that, a s though the numbers of bodies had actually declined in the older rabbit. DR. CUTZ: Yes. DR. McDOWELL: Do you feel confident in that, that they have actually declined, or is it a dilution effect? DR. CUTZ: Well, there is definitely a dilution effect because, you know, the volume of lung increases many times. But the added volume is in the alveolar region. If you look a t bronchial volume, there is minimal increase. So there is a huge increase in alveoli, but that’s not where the NEBs are. DR. McDOWELL: But did they just disappear then? You did serial sections, right? DR. CUTZ: Yes. Right. DR. McDOWELL: So are there any other studies that have been so careful. DR. CUTZ: Well, I think there is some species variation but, you know, the species we look mostly at is the human. I was just mentioning to Dr. Sorokin that the first reaction to the report of neuroepithelial bodies was amazement. People said, “We’ve looked a t thousands and thousands of sections. We’ve never seen it.” Some very eminent people. But they were looking a t adult lungs. If one looks a t the fetal or neonatal lung and then looks a t the adult lung side-by-side there is unquestionable prominence in one and paucity in the other. Now, it doesn’t mean that NEBs have totally disappeared. The other thing I’d like to say is that when one compares this phenomenon with nonpulmonary chemoreceptors such as carotid body or taste buds, the opposite happens. The carotid body starts small, then it grows and finally stabilizes to stay the same throughout life, the same as taste buds. So there is something special about neuroepithelial bodies. DR. HOYT: But Elizabeth and I have both shown you airway maps from Stephen Sarikas’s work in the hamster. He also did counts of neuroepithelial bodies during development. In the infracardiac lobe he counted them all and compared them day-by-day. They increase as you approach term, and yet there are many, many fewer a day before birth than there are in the infracardiac lobe in the adult, where we also counted all the cells and NEBs. DR. McDOWELL: Fewer in the babies than in the adult? DR. HOYT: Yes. In other words, in the hamster, where everything is rapid, the neuroendocrine cell population is blossoming and birth overtakes it before it’s fully established. I think that’s why one has to consider the timing of these things in the various species. Here is one species where it seems definitely that there are fewer late in gestation than there are in the adult. DR. McDOWELL: You say you’ve counted in the hamster, but how many people in the whole world have serially sectioned adult human lung? DR. HOYT: Years ago at Harvard, when I presented the hamster study at Pathology Grand Rounds-Lynne Reid said, “Well, now I’ll send you a n infant lung.” And I said, “I’m leaving town.” DR. McDOWELL: But the information just doesn’t exist. DR. HOYT: No, it doesn’t. So you have to look for indications. The problem is that the species are going to be different, and that one fact is a n enormous wild card thrown out into any attempt to make a major generalization on this point. DR. H U N G As far as the neuroepithelial bodies in the adult rabbits, some years ago we did a series of studies starting from 27 days gestation, 29, 30, and then 2 days postnatal onwards until adult, and our way WORKSHOP ON PULMONARY NEUROENDOCRINE CEJ,LS of calculating the number of neuroepithelial bodies was a little bit different from Dr. Cutz’s approach. We used glyoxylic acid-induced serotonin fluorescence. At the time a lot of people were using that, as it was a quick way of looking at those cells. What we calculated was the total number of neuroepithelial bodies taking into account the growth of the lung. And the data show that actually there is no reduction if you considerDR. CUTZ:-No reduction? DR. HUNG:-If you consider the total number in the entire lung from the 27 day fetus onward to-I think we did up to 5 months. DR. CUTZ: But it’s total lung volume. DR. HUNG: Right. It’s based on total lung volume. I realize that there is quite a bit of expansion due to growth of the alveolar region, but the fact is, though, the total number of neuroepithelial bodies is maintained. Now, one discrepancy probably is that there is a reduction of the size of individual neuroepithelial bodies. It dropped from about 40 microns diameter in late prenatalheonatals to less than half of this size in the adult. So, there possibly is either a reduction in cell numbers per neuroepithelial body or maybe a reduction of individual cell size. We don’t know which because using that fluorescent method one really cannot pick up individual cells that well, so we didn’t try to count the number of cells per neuroepithelial body. DR. SCHEUERMANN: In the neonatal rabbits? DR. HUNG: Yes. Those were rabbits. DR. SCHEUERMANN: But not, so far as I know, in other species, is there diminution of neuroendocrine cells. DR. HUNG: I have not seen anybody do that kind of calculation based on total lung. I don’t know if anybody has done that or not. DR. LINNOILA: I was just wondering why is i t so important to know whether the actual number is different in adults or not? That’sjust a numbers question. Is it worth investing say 10 people for 3 years to section the lungs and find out? Yes, it’s something we don’t know, but we don’t know the answer to many questions. DR. SPRINGALL: Well, maybe there is a n easier way of doing it than in whole lungs. There are other techniques you can use. DR. LINNOILA: Yes, but why is it important? What is the next question? To what kind of questions is t h a t question so important? DR. HOYT: I think this became a point of interest merely because neuroepithelial bodies could be identified in fetal lungs simply on the basis of their morphology. All you had to do was look carefully at sections. NEBs have a prominent organoid appearance. They lose their glycogen early. The surrounding epithelium is undifferentiated. It was easy to find them. They became less easy to find as birth approached and very difficult to find with relatively simple techniques in adult lungs. Based on those early observations, i t was thought that the cells had a role to play only during development or neonatal respiratory adaptation, and that if they persisted into adult life, they were vestigial rem- 67 nants that had no function. That’s why people became exercised over this particular issue. We don’t know whether there is a change in function from development to adult life. There may, in fact, be a waning, a loss, a degeneration, of some cell clusters that were formed early. Their places may be taken over by bodies formed anew in the lining epithelium. But we don’t know. That was how we got started with our continuous thymidine labeling. I had the idea that you could label all NEBs very heavily as they formed before birth, sacrifice animals for the next year, and then look at autoradiographs to see how long the cells persisted. Well, I got sidetracked and haven’t done that. DR. SCHEUERMANN: In my opinion the widely held view that neuroepithelial bodies involute shortly after birth is problematic. Recent quantitative studies, a t least on species investigated by Hung and his group (rabbits), the group of Sorokin and Hoyt (hamsters), Gosney and Sissons (rats), and Gosney, Sissons, and O’Malley (humans), indicate that the absolute number of neuroendocrine cells in adults almost equals that in late fetal life and that cell renewal continues throughout life. However, as the lung expands postnatally, many more nonendocrine than endocrine epithelial cells are added. That is my view of this problem. There is a dilution problem, and a decline in the number of neuroendocrine cells was never proved. DR. SOROKIN: Dr. Keith, did you have a comment? DR. KEITH: Yes. I agree with Ilona that the actual number of cells isn’t really that important in adult life, but the possibility that they are reserves that can kick in when there is a demand is important, and we should concentrate our work on what those demands are and what hormones are involved in that. DR. HOYT: Even if there are fewer cells, it all depends on their siting. I think they are placed a t critical points. One can draw a n analogy with the central nervous system. If you look a t the density of neurons you will find it much higher before birth and a t term than among us here today, and yet we still all manage. DR. SOROKIN: Well, perhaps then we can put it this way diplomatically, whether or not there is a postnatal involution remains a question, but even with considerable reduction in the adult, if these cells are needed, they can be induced to form. So they are a reality for serious investigation a t all periods, and the question no longer burns. DR. SCHEUERMANN: Perhaps it may really be a problem of expression of the stem cells to pneumocytes as well as to neuroendocrine cells. DR. CUTTITTA: Just one last question. Again, I’m really interested. Are there disease states, be it asthmatic or malignant states, or smokers, or what have you, that kick in increased numbers of NEBs for sure? DR. POLAK: Absolutely. In human disease, in many situations, not only in the lung, what’s happening during fetal development happens in the adult. When there is a n insult, a n injury, the cells are there for growth and repair, exactly a s we were discussing during development. And you’ll hear that today, if you stay. (Laughter.) DR. CUTTITTA: Then I have to stay. 68 WORKSHOP ON PULMONARY NEUROENDOCRINE CELLS Session 11. Neuroendocrine Cells in Adult Lung INTRODUCTION DR. POLAK: Ladies and gentlemen, we are ready to start the afternoon session. Before we commence I’d like on behalf of my colleagues to thank the organizers, Dr. Gail and Dr. Sorokin, very much for giving the European contingent in particular a n opportunity to share with you the work we are doing on the other side of the Atlantic. I would like to open this session by summarizing techniques that can be used, hopefully, to answer a number of outstanding questions. As we’ve heard today the lung is very well provided with what we call the diffuse neuroendocrine system. We have already had extensive discussion of the relationship between epithelial endocrine cells and the nervous system. In addition, the endothelium is now known to produce some of the same or similar regulatory peptides or factors. As you are very well aware, there are many ways to investigate the endocrine cells. Years ago toluidine blue was used extensively. Techniques to demonstrate the presence of secretory granules, or silver impregnation methods for the demonstration of endocrine differentiation, were also used. And of course, electron microscopy reveals the presence of electron-dense secretory granules. Several of the newer methods briefly discussed here have been applied to the lungs and are illustrated in the paper by Springall and Polak (these Proceedings, p . 96). Among these, immunocytochemistry is especially valuable because there are many markers available to pick up the endocrine cells, the nervous system, o r the endothelial cells. Using this technique, considerable morphological detail can be seen, even cell processes extending into the basement membrane. Markers for chromogranin, peptides, or other secretory granule components of course depend on the granule content of the endocrine cell in question, and if it is poorly granulated, the endocrine properties will be very difficult to demonstrate. Numerous antibodies are available for bombesin-like peptides, including several for the actual active segment, a s well as for the C-flanking terminal segment of the probombesin molecule. Endothelin is reasonably abundant in the lung; it exists in many forms (endothelin-1,2, and 3 predominating) and is not confined to endocrine cells. In the mouse lung it is distributed throughout the airway epithelium, but in man-especially in the healthy human lung and neonates-it immunolocalizes particularly to the endocrine cells, and of course also to the endothelium, where it is best demonstrated after use of enhancing techniques like glucose oxidase, diaminobenzidine, and nickel intensification. I will only touch on the potential use of confocal microscopy to quantify and reconstruct very heterogeneous tissue. This new microscope uses laser illumination and optically sections a rather thick layer of tissue like a bronchus or a whole lung. The information gained from each section is stored in a computer and then reconstructed, allowing us to see tortuous, threedimensional pathways that would be foggy or invisible under a conventional microscope. Dr. Springall has used confocal microscopy to demonstrate the entry of nerves into a neuroepithelial body immunostained for CGRP (see color plate, p . 170),and to visualize the entire plexus around pulmonary blood vessels. In addition, data obtained by this technique are quantifiable. Although in situ hybridization is targeted a t many fewer molecules than immunocytochemistry-since one molecule of message generates many molecules of protein product-it has proved a practicable method and is now fairly well established. The desired cDNA is inserted into a n appropriate vector that in the presence of polymerases will generate complementary RNA or messenger RNA for controls. These can be radioactively or nonradioactively labeled and then used to localize the messenger RNA molecules a t the microscopic level. In our studies on endothelin in fetal lung, in situ hybridization using complementary RNA molecules produced the same classical distribution for endothelin messenger RNA as the translated protein demonstrated by immunocytochemistry. If the signal of in situ hybridization is recorded on film, it can be quantified by densitometry and image analysis with use of radioactive tissue paste standards. It can also be quantified from emulsion-coated slides by counting grains in the cells, and when these are not all distributed in the same plane, analysis can be assisted by confocal microscopy. Another technique that can be combined together with immunocytochemistry and in situ hybridization is receptor technology and s t u d y of binding sites. In vitro autoradiography is one approach to finding bindings sites. This involves localizing binding sites in tissue sections incubated with a radioligand, controlled by competition studies with cold ligand, which will generate a negative image in autoradiograms. As with other applications of autoradiography, data can be recorded and analyzed from film or from emulsion-coated slides, where the silver grains indicate localization of the binding sites. So it’s possible to characterize the binding by Scatchard kinetics analysis for membrane preparations a s well as to have precise localization a t the microscopic level. A variant of this technique is to use in situ hybridization once the receptor has been cloned. Finally, the cell blot a s s a y offers the potential to quantitate hormone release combined with microscopical studies. This assay was first proposed by Kendall and Hymer (1989).It analyzes the secretion of individual cells and is useful a t both light and electron microscopic levels. One needs to settle dispersed cells on a translucent membrane able hold them as well as to bind peptides as they are released. It lends itself to experimental manipulation of the endocrine cells, inasmuch as stimulators or inhibitors can be applied, with responses measured by the magnitude of peptide secretion. After immunostaining with a n appropriate antibody, this is visualized as halos about the cells, and the data can then be made quantitative by image analysis. I hope that during the rest of the workshop, we may discuss further the use of such modern microscopical methods to understand the role of pulmonary endocrine cells in health and disease. (Applause.) DR. MILLER: I have one question regarding the re- WORKSHOP ON PULMONARY NEUROENDOCRINE CELLS 69 ceptor localization studies using radioactive ligands. DR. POLAK: Well, a lot of computer programs are How reliable do you think they are? I mean, you might now beginning to handle rotation, some of them very also have them bind to enzymes or binding proteins. In sophisticated. But before that there are a lot of probyour experience, has it been pretty reliable when you lems of resolution and other things to solve as wellcan correlate i t with in situ hybridization or immuno- penetration of the antibodies, the structure of the tishistochemistry? sue, how thick it is, and so on. It’s not a s simple a s I DR. POLAK: You are posing several problems there. made it look. The correlation is a very big problem. We need a comDR. SPRINGALL: Yes. It is a problem. To some explete workshop for it. If you look at neurotransmitters tent we’re still waiting on development of suitable obin the brain, you would expect to localize the binding jectives for confocal microscopy. They present a differsites where all the nerve terminals are. But there is ent set of problems from the normal light microscope. It what is called “receptor mismatch,” so the match is will happen, though. It’s definitely a technique which imperfect. The way you make sure is to apply a n enor- works best with very high resolution. It sounds odd. It’s mous number of controls for specificity, and still you converse to normal optical microscopy. You do have are only demonstrating “binding sites.” I don’t think problems with resolution, definitely, but they are being you can say “receptors” unless you can induce a n effect, overcome. And also the sort of three-dimensional mode.g., endothelin binding and induction of bronchocon- eling software that is obtainable now is largely restriction. So i t is better to say “binding sites.” That’s ceived from CAD designs, sort of commercial designing all. stuff. It hasn’t yet been properly adapted for scientific DR. SUNDAY: The cell blot assay, has that been use, but it’s becoming so, and that should help. used for looking a t directional secretion? DR. POLAK: For instance, it’s ideal for lung innerDR. POLAK: I haven’t seen anything on the cell blot vation. But of course all the problems we discussed this assay for directional secretion. It hasn’t been used a lot morning about the quantification of the endocrine cells and most of the work is in the pituitary. might be approached by staining and then reconstructDR. CUTTITTA: Julia, in that blot assay, have you ing, either with confocal or other new microscopical used different types of membranes? A multitude of techniques. membranes are now commercially available and some DR. SPRINGALL: The point is, I think, a t the mobind peptides far better than others. And relative val- ment the confocal microscope offers so much more that ues may be confused. Binding with nitrocellulose may it doesn’t matter if it’s not perfect yet. You can still get be low, but if you use a completely different membrane a lot of added information. it may be extremely high. NEUROENDOCRINE CELLS OF THE LUNG DR. POLAK: The characteristic of the membrane you need is one that will not disperse the cells but DR. SCHEUERMANN: Dear Dr. Polak, thank you keeps them grouped together (Cimini et al., 1993). very much for inviting me to the USA to give a lecture DR. SPRINGALL: I think Julia is right, that the on neuroendocrine cells on the occasion of this sympomost important thing is that cells can use i t as a sub- sium organized by the NIH, and I will also thank Dr. strate. If the cells float away, obviously you not only Gail for the possibility to give this lecture here. Thank lose your cells, but also you get no blots. you very much. (The following article is an expanded DR. STEPHENS: When you rotate these confocal im- version of Dr. Scheuermann’s talk.) ages, do you have a problem with resolution in the third dimension?