Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Lysozyme (LZ) is a low molecular-weight enzyme (15 kD) with bactericidal activity. It catalyzes the hydrolysis of -glycosidic bonds of peptidoglycan, an important component of the bacterial cell wall. It is secreted by macrophages and by many different exocrine glands and tissues, and is considered to be an important nonimmunologic antibacterial defense (for review, see 1, 2). LZ is also synthesized by epidermal keratinocytes, and the enzyme levels are increased in psoriatic skin lesions (3, 4). LZ genes of several species, including chicken and human, have been cloned and sequenced, and transcripts 1.6 and 0.6 kb in size have been detected in several human cell lines (for review, see 5, 6).

In the conducting airways, LZ has been localized immunohistochemically to the serous cells of submucosal glands (7, 8). However, biochemical studies have shown LZ also in the surface epithelium (9). In the pulmonary parenchyma, type II alveolar cells have been shown to secrete LZ together with lamellar body surfactant (10).

Studies in our laboratory are concerned with the regulation of differentiation of normal human tracheobronchial epithelial (NHTBE) cells by retinoids, particularly retinoic acid (RA) and epidermal growth factor, and the effects these bioregulators have on mucin and nonmucin secretions (11, 12). We showed (11) that NHTBE cells grown in the presence of RA in air-liquid interface (ALI) cultures secrete large amounts of mucin as well as small amounts of LZ, lactoferrin, and the secretory leukocyte protease inhibitor. Such cultures exhibit the morphologic features characteristic of columnar mucociliary epithelium. However, when RA is deleted from the medium, a squamous epithelium develops instead and mucin secretion ceases (11). This is reminiscent of many in vivo studies (for review, see 17) that have shown that deletion of vitamin A from the diet leads to extensive squamous metaplasia in the conducting airways. Thus, both the in vivo and in vitro studies indicate the retinoid dependency of normal mucociliary differentiation. One of the more surprising findings in our in vitro studies was that LZ secretion was greatly elevated in the RA-deficient cultures (11), suggesting that RA is a negative regulator of LZ. Paradoxically, LZ messenger RNA (mRNA) levels were decreased in RA-deficient compared with RA-sufficient cultures (16).

In the investigations described here, we wanted to determine when during the course of squamous differentiation LZ levels increase in RA-deficient NHTBE cultures. We also hoped to answer the question, raised in the previous study, whether RA directly regulates LZ levels.

Our studies suggested that the RA-deficient epithelium begins to accumulate intracellular LZ during early phases of squamous differentiation. As intracellular LZ increases to very high levels, cells die and release LZ onto the apical culture surface. Our findings indicate that the buildup of LZ in the squamous epithelium was not the result of increased LZ synthesis, but rather was due to decreased LZ catabolism. Conceivably, similar changes in LZ metabolism may also occur in vitamin A deficiency-induced metaplastic squamous differentiation in vivo.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

ALI Cultures

Passage-1 NHTBE cultures (strain 2002), purchased from Clonetics Corp. (San Diego, CA), were subcultured once and stored in liquid nitrogen (passage-2). Passage-2 cells were seeded onto type I collagen gel (Collaborative Research, New Bedford, MA)-coated, semipermeable membranes (Trans-clear; Costar Corp., Cambridge, MA) in serum-free, hormone- and growth factor-supplemented media (for a detailed description of cell culture methods, see 11). Cultures were grown submerged for the first 7 d, at which time an ALI was created. Cultures were maintained at 37°C in a humidified atmosphere of 5% CO₂ in air.

Passage-1, normal human epidermal keratinocytes (NH-EK) were obtained from Clonetics Corp. and expanded once on plastic tissue-culture dishes in fully supplemented keratinocyte growth media (KGM; Clonetics Corp.). ALI cultures of NHEK cells were grown under conditions similar to those described above for NHTBE cells. Passage-2 NHEK were seeded onto semipermeable membranes and grown in KGM. Cultures were maintained at 37°C in a humidified atmosphere of 5% CO₂ in air. Culture media were changed every other day. Apical secretions and cell extracts were collected at different times of culture, as required by the experimental protocol.

Immunodetection and Quantitation of LZ and Mucin Production

Methods for detection of mucin and LZ produced by cultured NHTBE cells using an immunoblot assay have been previously described in detail (11). Apical secretions accumulating over a 24-h period were collected and centrifuged to remove cellular debris. To determine the intracellular levels of LZ from NHEK and NHTBE cultures, cell lysates were prepared from cell suspensions that had been pooled from two or three cultures. Following enzymatic dissociation, the cells were pelleted, washed, and lysed using 50 mM Tris-buffered saline containing 1% Tween-20 (Bio-Rad, Hercules, CA), 1 mM ethylenediaminetetraacetic acid (EDTA), and protease inhibitors (Complete; Boehringer Mannheim, Indianapolis, IN), and briefly sonicated. Purified mucin (0.85 mg dry weight of purified mucin per milliliter, a generous gift from Dr. C. W. Davis, University of North Carolina, Chapel Hill, NC) and LZ (Sigma, St. Louis, MO) were used as standards. Mucin was detected using the monoclonal antimucin antibody 17Q2 (18) (a generous gift from Dr. Judith St. George, Genzyme Corp., Framingham, MA), and LZ was detected using the polyclonal antihuman LZ antibody (DAKO, Carpinteria, CA). Dilutions of apical secretions or cell lysates and standards were applied to nitrocellulose membranes and incubated with the appropriate primary antibody, followed by reaction with horseradish peroxidase (HRP)-conjugated goat antimouse or antirabbit immunoglobulin (Ig)G. The signal was detected by chemiluminescence (ECL kit; Amersham, Buckinghamshire, UK), and a standard curve was generated by linear regression analysis from which the concentration of individual samples could be determined. The specificity of the human anti-LZ antibody was confirmed by Western blot analysis. The number of cells per culture was determined by visual cell counts on dissociated cell suspensions following enzymatic treatment of the cultures. The data (representive of two or three repeat experiments) are expressed as the means ± SD of triplicate cultures from the same experiment. Statistical comparisons were made using Student's t test.

Metabolic Labeling and Immunoprecipitation

To measure the rate of LZ synthesis, Day-10 NHTBE RA-sufficient and RA-deficient cultures were incubated for 2 h with cysteine-methionine-depleted culture medium, which was then replaced for the indicated time periods with fresh medium containing ³⁵S-labeled methionine and cysteine (50 µCi/ml, L-[Pro-mix (³⁵S)] cell labeling mix; Amersham). To measure the LZ half-life, the ³⁵S-containing medium was removed after 4 h and replaced with fresh medium containing 3× cysteine and methionine (12 mg/ml L-cysteine and 24 mg/ml L-methionine). In both protocols, cell lysates were collected at various times. Cultures were first washed with phosphate-buffered saline and the cells were lysed in 0.25 ml of lysis buffer (50 mM Tris-HCl, pH 7.5; 150 mM NaCl; 1% Triton X-100; 2 mM EDTA; 1 mM leupeptin; and 1 mM phenylmethylsulfonyl fluoride [PMSF]). Cell lysates were homogenized by brief sonication and centrifuged (30 min at 4°C at 16,000 × g). The protein concentration was determined (Bio-Rad) and 1 mg protein lysate was incubated overnight at 4°C with 10 µg antilysozyme antibody (DAKO) in a 1-ml reaction volume. The immunocomplex was captured by binding to protein-A agarose beads (10 µg/ml reaction; Pharmacia, Piscataway, NJ) and pelleted. Immunoprecipitates were first washed in buffer A (50 mM Tris-HCl, pH 7.5; 150 mM NaCl; 0.5% Triton X-100; 2 mM EDTA; 1 mM leupeptin; and 1 mM PMSF), and then in buffer B (10 mM Tris-HCl, pH 7.5, and 0.1% Triton X-100). The immunoprecipitates were eluted in 50 µl of 2× Laemmli sample buffer by boiling for 5 min, and then analyzed by sodium dodecyl sulfate-polyacrylamide gel electrophoresis (SDS-PAGE) (10%). The electrophoresed proteins were transferred to nitrocellulose membranes, and signals from radiolabeled LZ were captured using a Storm PhosphorImager (Molecular Dynamics, Sunnyvale, CA). The radiolabeled lysozyme signal was quantitated using the associated analysis software (ImageQuant), and the data are expressed as arbitrary PhosphorImager units.

Northern Blot Analysis for Cornifin mRNA

Total RNA was isolated from NHTBE cell cultures using Tri-Reagent (Molecular Research Center, Cincinnati, OH), and 10 µg of RNA was fractionated by electrophoresis in a 1.5% agarose gel containing 6.6% formaldehyde and transferred to Nytran membranes (Schleicher & Schuell, Keene, NH) by capillary blotting. A full-length rabbit cornifin complementary DNA (cDNA) probe (19) (a generous gift from Dr. A. M. Jetten, NIEHS, Research Triangle Park, NC) was radiolabeled with ³²P-deoxycytidine 5'-triphosphate. Northern blots were hybridized (Quik-Hyb; Strategene, La Jolla, CA) for 1 h at 62°C and washed twice for 15 min at room temperature in 2× saline sodium citrate (SSC)/0.1% SDS, once for 30 min at 60°C in 0.1× SSC/0.1% SDS, and then briefly in the latter solution at room temperature. The signal of 28S ribosomal RNA (rRNA), which was used to normalize for RNA loading and integrity, was detected using ³²P end-radiolabeled oligonucleotide probe (Life Technologies, Gaithersburg, MD). The signals were detected using Hyperfilm (Amersham).

Reverse Transcription-Polymerase Chain Reaction (RT-PCR) for LZ mRNA

Because the LZ mRNA levels were often too low to be detected by Northern blot analysis, we used RT-PCR for the detection of LZ mRNA (16). Oligonucleotide primers were designed according to the published sequences for human LZ (20) (Genbank accession no. J03801, 5' primer: CTCTCATTGTTCTGGGGC; 3' primer: ACGGACAACCCTCTTTGC), which generated a 350-base pair (bp) fragment. Oligonuleotide amplimers for 2 microglobulin (2M, which was used as a control gene for RT-PCR) were purchased from Clontech Laboratories (Palo Alto, CA); they generated a 335-bp PCR fragment. RT-PCRs were performed using a Perkin-Elmer Cetus DNA Thermal Cycler according to the manufacturer's recommendations. Total RNA (1 µg/20 µl reaction volume) was reverse transcribed into cDNA using random hexanucleotide primers and Moloney murine leukemia virus reverse transcriptase. Forty percent of the RT reaction for LZ or 4% for 2M of the resulting cDNA was amplified using 0.2 mM of each primer. The optimized concentration of MgCl₂ in the PCR was 1.5 mM for both LZ and 2M. Denaturation was carried out at 95°C for 1 min. Annealing temperatures were 55°C for LZ and 60°C for 2M for 1 min; extension was performed at 72°C for 1 min.

We used comparative kinetic analysis to compare mRNA levels for each gene for each set of culture conditions as described previously (16). PCR products were separated by electrophoresis on a 2% Seakem agarose gel (FMC, Rockland, ME) containing 50 ng/ml ethidium bromide, and photographed with Polaroid Type 55 film. The negatives were scanned on a Molecular Dynamics Densitometer, and the signal was analyzed using ImageQuant software. The linear range for the PCR was established by plotting the intensity of signal versus PCR cycle number. The linear range for LZ was found to be between 25 and 30 cycles, and for 2M between 26 and 31 cycles. To verify that the amplified products were from mRNA and not genomic DNA contamination, negative controls were performed by omitting reverse transcriptase from the RT reaction. In the absence of reverse transcriptase no PCR products were observed. Specific amplification of LZ mRNAs was confirmed by the sequencing (dsDNA Cycle Sequencing System; Life Technologies) of PCR fragments.

Histology and Immunocytochemistry

NHTBE cells were grown in RA-sufficient or RA-deficient medium for 14 d. Intact cultures or cell pellets prepared from enzymatically dissociated cultures were fixed in 10% neutral buffered formalin, embedded in paraffin, and sectioned. Sections either were stained with hematoxylin and eosin (H&E) or were used for immunocytochemistry after deparaffinization and rehydration. LZ-positive cells were detected using a rabbit polyclonal antihuman LZ antibody (1:200 dilution; DAKO); cornifin-positive cells were detected using a rabbit polyclonal anticornifin antibody (1:1,000 dilution; a generous gift from Dr. Anton Jetten). The reactivity was detected by avidin-biotin binding using a commercially available kit (Vector Labs, Inc., Burlingame, CA). The specificity of the antibody reactions for LZ and cornifin was confirmed using normal rabbit IgG (Vector), following which no reactivity was detected. Proliferating cell nuclear antigen (PCNA) was detected by methods previously described (21). Briefly, slides made from cell pellets were first incubated for 30 min at 37°C in 2.0 N HCl, then washed and reacted with the mouse monoclonal anti-PCNA antibody 19A2 (1:100 dilution; Coulter Immunology, Hialeah, FL). PCNA reactivity was detected by streptavidin immunolabeling technique using a biotinylated goat antimouse IgM antibody (1:400 dilution; Jackson Immunoresearch Labs, West Grove, PA) and streptavidin-conjugated HRP-labeled complex (Biogenex Laboratories, San Ramon, CA). Mouse anti-IgG served as the control antibody. Diaminobenzidine was used as the chromagen to detect reactivity with all antibodies.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Time Course of Differentiation and LZ Expression

To determine when abnormal LZ expression occurred during squamous differentiation of RA-deprived NHTBE cultures, cells were grown in RA-free medium and, for comparison purposes, in RA-supplemented medium. Mucin secretion was used as a marker of mucous differentiation, and cornifin mRNA expression as an indicator of the squamous phenotype (19, 22, 23). In the same cultures, intracellular and extracellular LZ protein and mRNA were measured. As seen in Figure 1A, RA-deficient cultures secreted no or only insignificant amounts of mucin (1 µg/10⁶ cells or less); and as a sign of squamous differentiation, cornifin mRNA was strongly expressed at all time points (Figure 1B). In contrast, RA-sufficient cultures secreted nearly 10 µg of mucin per 10⁶ cells on Day 7 and between 100 and 500 µg of mucin per 10⁶ cells from Day 10 on, indicating that the mucous phenotype was fully developed. Cornifin levels were low in these cultures.

View larger version (54K): [in this window] [in a new window]   Figure 1.   Mucin secretion and cornifin expression during different phases of differentiation of RA-sufficient and RA-deficient NHTBE cultures. NHTBE cells were grown in RA-sufficient ( solid columns) or RA-deficient (hatched columns) medium. (A) Apical secretions were collected and assayed for mucin. Data represent the means ± SD of triplicate cultures. (B) Total RNA was collected at the indicated times, and the levels of cornifin mRNA and 28S rRNA expression were determined by Northern blotting.

The morphologic features of the cultures reflected the state of differentiation suggested by the mucin and cornifin measurements. On Day 7 the epithelium of RA-deficient cultures was beginning to stratify (Figure 2A) and 16% of the cells reacted with cornifin antibody. On Days 10 and 14 the epithelium had become multilayered and squamous (Figures 2B and 2C); cornifin staining was most intense in the suprabasal cell layers (Figure 2D), and 80 to 90% of the cells reacted with cornifin antibody (Figure 2E). RA-sufficient cultures contained no cornifin-reactive cells (Figure 2F).

View larger version (110K): [in this window] [in a new window]   Figure 2.   Morphologic and immunocytochemical manifestations of squamous differentiation in RA-deficient NHTBE cultures. Histologic sections of RA-deficient cultures: (A) Day 7, H&E; (B) Day 10, H&E; (C ) Day 14, H&E; (D) Day 14, LZ antibody staining (arrowheads indicate positively stained suprabasal cell layers). Cornifin immunocytochemistry on cell pellets obtained from Day-14 RA-deficient (E) and RA-sufficient (F ) cultures. No staining was detected in cell pellets from RA-deficient cultures in which the anticornifin antibody was replaced with equivalent amounts of normal rabbit IgG. Bar: 50 µm.

LZ measurements (Figure 3A) showed that intracellular LZ increased in the RA-deficient cultures from 1.3 µg/ 10⁶ cells on Day 7 to 5.0-9.5 µg/10⁶ cells on Days 16 through 18. In comparison, intracellular LZ levels were low in RA-sufficient cultures, fluctuating between 0.3 and 0.6 µg/10⁶ cells throughout the time period studied. Extracellular LZ became detectable for the first time on Day 12 in both sets of cultures (Figure 3B). In RA-deficient cultures it increased from 0.08 µg/10⁶ cells on Day 12 to 5.0-7.5 µg/10⁶ cells between Days 16 and 18. In RA-sufficient cultures it rose from 0.03 µg/10⁶ cells on Day 12 to 0.3 µg/10⁶ cells on Day 18. LZ mRNA levels were low on Day 7 in both RA-deficient and RA-sufficient cultures (Figure 3C). Between Days 10 and 14, LZ mRNA was strongly expressed in both sets of cultures. Surprisingly, on Days 16 and 18, LZ mRNA levels were much lower in RA-deficient than in RA-sufficient cultures, despite the fact that the former contained much higher levels of LZ protein than did the latter. Similar LZ protein and mRNA expression patterns were seen in other independent time-course studies. Expression of the control gene 2M was constant throughout.

View larger version (35K): [in this window] [in a new window]   Figure 3.   Intra- and extracellular LZ levels and LZ mRNA expression during different phases of differentiation of RA-sufficient and RA-deficient NHTBE cultures (same experiment as in Figure 1). (A) Cell lysates were collected and pooled from two or three cultures at each time point, and the levels of intracellular LZ protein were determined. (B) Extracellular LZ protein present in the apical washings; data represent the means ± SD of triplicate cultures. (C ) LZ and 2M mRNA levels were determined by RT-PCR from total RNA.

In previous experiments we had noticed that cell exfoliation was much more conspicuous in RA-deficient than in RA-sufficient cultures. We therefore estimated the number of cells exfoliating in the two sets of cultures. As shown in Figure 4A, in the RA-deficient cultures the rate of exfoliation was already > 3 × 10⁵ cells/24 h on Day 12 and increased to > 1 × 10⁶ cells/24 h on Day 16. (In Day-14 RA-deficient cultures the amount of LZ was 24.9 µg/ 10⁶ exfoliated cells, compared with 8.9 µg/10⁶ attached cells.) Despite the rapid cell loss, the cultures remained in a steady state (Figure 4B) because increased cell proliferation compensated for the increased cell death. The proportion of cells expressing PCNA, a marker for cycling cells (21), was 15% on Day 14 in RA-deficient cultures compared with < 1% in RA-sufficient cultures.

View larger version (25K): [in this window] [in a new window]   Figure 4.   Growth and cell exfoliation in RA-sufficient and RA-deficient cell cultures. (A) Cells exfoliating over a 24-h time period were collected at each time point. The samples of three cultures were pooled, and the number of cells was determined in RA-sufficient ( solid columns) and RA-deficient (hatched columns) cultures. (B) The number of attached cells was determined following enzymatic dissociation of RA-sufficient ( solid square) and RA-deficient (open circle) cultures. The data for attached cells represent the mean cell numbers ± SD obtained from triplicate cultures.

LZ Expression in Epidermal Keratinocytes

Epidermal keratinocytes are known to synthesize low levels of LZ (3, 4). To determine whether intracellular LZ accumulation occurred normally during squamous differentiation or whether this was a feature of abnormal squamous differentiation of vitamin A-deficient airway epithelium, we compared the levels of LZ in cultures of fully differentiated NHEK with those in metaplastic, squamous NHTBE cultures. Both sets of cultures were grown in RA-free medium. NHEK cells were grown until confluence in medium containing low levels of Ca²⁺, and terminal differentiation was induced by switching the cultures on Day 5 to medium containing 2.0 mM Ca²⁺. Apical washings and cell lysates were collected from Day-13 cultures. As seen in Figure 5, the level of extracellular LZ present in the apical washing of NHEK cultures was 0.09 µg/10⁶ cells, whereas the apical washing from NHTBE cultures contained 1.75 µg/10⁶ cells. Intracellular LZ levels of NHEK cultures were 0.45 µg/10⁶ cells compared with 5.25 µg/10⁶ cells in NHTBE cultures. These findings indicate that accumulation of high levels of LZ does not occur during normal squamous differentiation of epidermal cells but seems to be a feature unique to abnormal squamous differentiation of retinoid-deficient airway epithelial cells.

View larger version (50K): [in this window] [in a new window]   Figure 5.   Comparison of LZ levels in NHTBE and NHEK cultures grown in the absence of RA. NHTBE (hatched columns) and NHEK ( solid columns) cells were grown in RA-free culture media. The extracellular and intracellular levels of LZ were determined in late-stage cultures undergoing squamous differentiation.

LZ Synthesis and Degradation

We considered two possible mechanisms that might be responsible for the increased accumulation of LZ in RA- deficient cultures, namely, either increased LZ synthesis or decreased LZ degradation. To examine the kinetics of LZ protein synthesis, Day-10 RA-deficient and RA-sufficient cultures were incubated over a 6-h period with ³⁵S- methionine-cysteine and cell lysates were collected first at 0.5 h and then at hourly intervals. The newly synthesized LZ was immunoprecipitated from the lysate with anti-LZ antibody and separated by SDS-PAGE. We chose Day 10 of culture for these studies because, as shown in Figures 3 and 4, at that time intracellular LZ levels were already much higher in RA-deficient than in RA-sufficient cultures; however, extracellular levels were still undetectable and the rate of cell exfoliation was still low. Therefore, synthesis and degradation of LZ could be monitored without interference from LZ secretion or release. As shown in Figure 6, the levels of labeled LZ were higher in RA-sufficient cultures at all time points than in RA-deficient cultures. However, because uptake of radiolabeled precursors turned out to be significantly greater in RA-sufficient than in RA-deficient cultures (data not shown), it was necessary to normalize for these differences in uptake kinetics. After making this adjustment, we found that LZ synthesis in RA-sufficient cultures was still marginally higher than in RA-deficient cultures. Therefore, these findings ruled out the possibility that the high levels of LZ protein found in RA-deficient cultures were due to increased LZ synthesis. On the contrary, the rate of LZ synthesis was reduced in the RA-deficient cultures and so was the rate of overall protein synthesis (data not shown). The data presented in Figure 6 show the unadjusted incorporation of labeled precursor into LZ protein.

View larger version (24K): [in this window] [in a new window]   Figure 6.   LZ protein synthesis in RA-deficient and RA-sufficient cultures. Day-10 RA-deficient (hatched columns) and RA-sufficient ( solid columns) cultures were incubated with 35S-methionine- cysteine for 6 h, and LZ synthesis was detected at the indicated times by immunoprecipitation (inset) and quantitated using PhosphorImager and ImageQuant analysis.

To determine whether the decay rate of intracellular LZ was different in RA-sufficient than in RA-deficient cultures, Day-10 cultures grown in the presence or absence of RA were labeled with ³⁵S-methionine-cysteine for 4 h and the labeling media were then removed and replaced with medium containing excess cold methionine-cysteine. Cell extracts were collected over the following 8 h and immunoprecipitated with anti-LZ antibody, and the immunoprecipitates were separated by SDS-PAGE. Figure 7 shows that in the RA-deficient cultures, the levels of newly synthesized LZ barely changed over the 8-h post-labeling period. In contrast, in RA-sufficient cultures, in which the levels of newly synthesized LZ were much higher at the end of the labeling period than in the RA-deficient cultures (see Figure 6), radiolabeled LZ levels decreased rapidly during the 8-h post-labeling period, with a half-life of approximately 6 h. The data thus suggest an increased half-life for LZ in RA-deficient cultures. This conclusion was supported by the results of studies in which the stability of LZ protein in 10-d-old RA-sufficient and RA-deficient cultures was measured after blocking protein synthesis with cycloheximide (data not shown).

View larger version (24K): [in this window] [in a new window]   Figure 7.   Degradation of intracellular LZ protein in RA-deficient and RA-sufficient cultures. Day-10 RA-deficient (hatched columns) and RA-sufficient ( solid columns) cultures were prelabeled for 4 h with 35S-methionine-cysteine. Radiolabeled LZ was detected from immunoprecipitates (inset) collected at the indicated time intervals over an 8-h period, and the levels were quantitated as in Figure 6.

Effects of RA Treatment on LZ Expression in RA-Deficient Cultures

The experiments summarized in Figure 3 raise the question whether intracellular lysozyme levels were regulated directly by RA or indirectly as a consequence of RA-dependent differentiation. We argued that if LZ were directly regulated by RA, LZ mRNA or protein levels should change rapidly following RA treatment. To address this question, RA-deficient cultures were treated on Day 7 with 10⁶ M RA and the effects of the treatment on LZ protein and mRNA levels were examined at 24-h intervals. As shown in Figure 8A, intracellular LZ protein levels were not significantly reduced until 3 d after addition of RA to the cultures; thereafter, they remained low (between 0.2 and 0.25 µg/10⁶ cells). In comparison, LZ protein levels increased in untreated cultures from ~ 1.5 µg/10⁶ cells at the beginning of the experiment to > 20 µg/10⁶ cells 7 d later. Interestingly, LZ mRNA levels (Figure 8B) were higher in RA-treated than in untreated cultures at all time points, except at 24 h after the start of treatment. The mRNA levels of the control gene 2M were not significantly affected by RA treatment. Mucin secretion was significantly increased 3 d after the start of RA treatment and continued to increase subsequently, as a sign that the mucous phenotype had been restored (Figure 8C). Untreated cultures made either little or no mucin.

View larger version (32K): [in this window] [in a new window]   Figure 8.   The effect of RA treatment on intracellular LZ protein and LZ mRNA levels of RA-deficient cultures. NHTBE cells were grown for 7 d in RA-deficient medium. Cultures either were maintained for an additional 7 d in RA-deficient media or were treated with 106 M RA. At time 0 and at daily intervals, the levels of intracellular LZ protein (A) were determined in cell lysates pooled from two or three cultures per time point; the levels of LZ and 2M mRNA (B) were determined by RT-PCR from total RNA, and the amount of mucin secreted (C ) was determined from apical washings (data represent the means ± SD from triplicate cultures).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

The molecular mechanisms of the regulation of LZ gene expression have been extensively investigated and have been found to be highly species- and cell type-specific. (24-26; for review, see 2). Recently, the 5'-flanking region of the bovine LZ gene, lys 5a, has been shown to contain binding sites for serous-cell nuclear proteins and serous cell-specific transcriptional activity (27). Several cis-acting promoter and enhancer regions of the chicken LZ gene have been isolated and characterized (2). The two most commonly used models to study the regulation of LZ gene expression are the chicken oviduct and mouse macrophages. In the chicken oviduct, LZ expression is steroid hormone- dependent, whereas in the monocyte/macrophage pathway of differentiation, LZ expression has been closely linked to macrophage maturation and to activation of macrophages by immune regulators (6).

Very little is known about regulation of LZ in the respiratory tract beyond the fact that secretagogues and neutrophil elastase can stimulate its secretion (8, 28). We previously found that RA-deprived NHTBE cultures that undergo metaplastic squamous differentiation secrete or release increased amounts of LZ, whereas mucin secretion is drastically reduced compared with RA-sufficient cultures (11, 16). RA-sufficient cultures, on the other hand, secrete large amounts of mucin but only small amounts of LZ, thus suggesting that RA downregulates LZ secretion and upregulates mucin secretion. It was our intention to explore further the relationship between RA deficiency- induced squamous differentiation and the previously reported changes in LZ secretion. Our studies showed that LZ accumulated intracellularly in RA-deficient cultures as squamous differentiation progressed, and large numbers of superficial cells containing high levels of LZ exfoliated, releasing LZ at the apical culture surface. Thus LZ, instead of being actively secreted as we had previously inferred (11, 16), appeared to be released during exfoliation of the superficial cell layer. The total amount of LZ (intra- and extracellular LZ combined) was much greater in RA-deficient than in RA-sufficient cultures.

To shed light on the mechanisms that led to the excessive accumulation of LZ in the metaplastic cultures, we examined LZ synthesis and degradation in RA-deficient cultures at a time (Day 10 of culture) when the intracellular levels were clearly elevated but before surface cells started to exfoliate in great numbers and before LZ was released extracellularly. Contrary to our expectation, we found that LZ synthesis was not increased in RA-deficient cultures. In fact, the rate of LZ synthesis was slightly decreased over that in RA-sufficient cultures. (So was overall protein synthesis, even after correcting for differences in labeled precursor uptake; data not shown.) In addition, the decay rate of the newly synthesized LZ was much slower in RA-deficient than in RA-sufficient cultures. This explains why the RA-deficient cultures contained so much more LZ than did the RA-sufficient cultures, despite the fact that their rate of LZ synthesis was not elevated. It seemed that the metaplastic NHTBE cells not only had lost the secretory mechanism available to normally differentiated airway cells to rid themselves of LZ, but also had a much- reduced ability to degrade LZ. That this was a feature of metaplastic rather than normal squamous differentiation was suggested by the finding that normal human keratinocytes did not accumulate high levels of LZ during squamous differentiation.

It was of interest to determine whether the effect of RA on LZ was a direct effect on LZ transcription or translation, or whether it was an indirect consequence of the regulation of epithelial cell differentiation by RA. We found that RA treatment of RA-deficient cultures did not reduce steady-state levels of LZ mRNA; on the contrary, LZ mRNA levels were modestly increased at all but the earliest time points after the start of treatment. LZ protein levels did not decrease until 3 d after the start of treatment, which coincided with a major increase in mucin secretion (i.e., with restoration of the mucous phenotype). These findings suggest that the RA effects on LZ were probably indirect, secondary to its control of airway cell differentiation. Our studies showed, however, that LZ mRNA levels were regulated by two other factors. With the onset of differentiation (or beginning of confluence) between Days 7 and 10 of culture, LZ transcript levels increased markedly in both RA-sufficient and RA-deficient cultures. In the late RA-deficient cultures, in which the intracellular LZ concentrations were very high, LZ mRNA levels decreased substantially, perhaps indicating that a negative feedback mechnism existed between LZ protein and mRNA.

Since the 1925 report by Wolbach and Howe that vitamin A deficiency induced squamous metaplasias in rats (29), many investigators have studied this aberrant form of differentiation of airway epithelium. In more recent studies, McDowell and colleagues (17) examined in great detail the time-dependent development of metaplastic squamous differentiation in the respiratory tract of vitamin A-deficient hamsters. These investigators reported high levels of cell proliferation in the basal cell layer of advanced squamous metaplasias and noted increased cell sloughing from the surfaces of such lesions. They suggested that growth factors and cytokines released by inflammatory cells infiltrating the airway epithelium might be causing the increased cell proliferation. We also observed increased cell turnover in late stages of metaplastic differentiation in RA-deficient NHTBE cultures. Surface cells were dying at an increased rate but were replaced by rapid cell proliferation, as indicated by high PCNA counts, thus maintaining the cultures in a steady state. We propose that the accelerated cell death of LZ-laden superficial cell layers was responsible for driving the compensatory increase in cell proliferation. It is conceivable that the high levels of intracellular LZ, a highly basic protein with enzymatic activity, might damage the cells and lead to premature cell death. Such a mechanism might also be involved in vivo, causing increased cell sloughing and compensatory cell proliferation of metaplastic airway epithelium in vitamin A-deficient animals.

In conclusion, we propose that LZ accumulation in RA-deprived NHTBE cultures is the result of a defective LZ metabolism and is part of the epithelial pathology of RA deficiency-induced metaplastic airway epithelium.