Loss of GFI1 impairs pulmonary neuroendorine cell proliferation, but the neuroendocrine phenotype has limited impact on post-naphthalene airway repair

Animals and Treatment

Wild-type (wt) and Gfi1^−/− mice on a C57Bl/6 background were used at 2-months of age. Groups of wt and Gf1 null littermates were utilized throughout the study. Naphthalene (Sigma, St Louis, MO, USA) was dissolved in corn oil and administered intraperitoneally as described previously.¹¹ Wt and Gfi1^−/− control mice were injected similarly with corn oil alone. The mice were killed 3 or 5 days following treatment. Two hours before death, bromo-deoxyuridine (BrdU) (Sigma) was administered. All control and experimental animals received BrdU. BrdU was dissolved in water at 10 mg/ml and administered intraperitoneally at 10 μg/g body weight.

Lung Fixation and Tissue Collection

Mice were killed, and the lungs were harvested after formalin fixation as previously described with minor modifications.¹¹ Right lung lobes were fixed by inflation at 10 cm H₂O with neutral-buffered Formalin instilled through a tracheal cannula.¹¹ The harvested lungs were further fixed in 10% neutral-buffered formalin (Surgipath, Richmond, IL, USA) overnight at room temperature. The 10% neutral buffered formalin was replaced with 70% ethanol and the tissues were embedded into paraffin.

Immunohistochemistry

Paraffin-embedded tissue sections were deparaffinized, hydrated and stained using the Vectastain ABC kit (Vector Laboratories, Burlingame, CA, USA) following the vendor’s instructions with modifications as described.³² Rabbit anti-CGRP (1:3000, Amersham Biosciences, Piscataway, NJ, USA) and rabbit anti SYN (1:100, Zymed, San Fransisco, CA, USA) antibodies were employed to evaluate NE differentiation, whereas rabbit anti-Clara cell 10 kDa secretory protein (CC10) (1:100 000, Dr Franco DeMayo, Baylor School of Medicine, Houston, TX, USA) was used to identify the major secretory product of the Clara cell, a signature of epithelial cell differentiation within the conducting airways. A rat anti-BrdU antibody (1:300, Accurate Chemical, Westbury, NY, USA) was used to detect BrdU incorporation in cells undergoing proliferation. Following immunohistochemical staining NE differentiation was quantified by counting the positive foci composed of solitary PNEC and NEBs containing CGRP or SYN in airways. NE expression was computed for each animal by dividing the number of positive NE foci (solitary PNEC or NEBs) by the total number of airways analyzed (Foci/Airway Ratio or F/A). CC10 immunostaining pattern was assessed using a staining index as described previously,³³ which combines the intensity and distribution of positive cells.

Morphometric Analysis of Proliferation and Apoptosis

Apoptotic index was determined by the TUNEL assay using the Apoptag Plus Peroxidase in situ Detection Kit (Chemicon International, Inc., Temecula, CA, USA). according to the vendor’s instructions. Quantification of BrdU incorporation and TUNEL staining was assessed using MetaMorph imaging software (Universal Imaging, Westchester, PA, USA) with a ×20 objective (Nikon Eclipse E400 Microscope). A labeling index (LI) for airway epithelial cells was derived by dividing the total number of positive nuclei by 1000, the total number of cells analyzed per animal.

In addition to the overall labeling index, terminal bronchioles were defined as described previously,¹⁸ and cells located 200 μm from the bronchioloalveolar junction were analyzed. To determine the rate of BrdU incorporation in NE cells adjacent sections placed on the same slide were stained for CGRP and BrdU and the percentage of labeled PNEC and NEBs was established.

Statistical Analysis

Mean values for all parameters were compared with control values using the Mann–Whitney Rank sum test unless indicated otherwise. Statistical significance for all comparisons was acknowledged when P-values were <0.05. The comparison of the fold-change between wt and Gfi1^−/− were determined using two-way ANOVA. The factors were treatment (naphthalene vs oil) and type (wt vs Gfi1^−/− mice). A variance stabilizing transform function was used for the percentages (the arcsin square root). With this model, a significant interaction would mean that the effect of treatment (naphthalene vs oil) will be different depending on type (wt vs Gfi1^−/−). This model does not test whether the ratios are different between the two types, but tests whether the magnitude of the effect measured on the transformed scale (arcsin square root of percent) are different between the two types. The SPSS statistical software (SPSS Inc. Chicago, IL) was used for performing the two-way ANOVA analysis.

Differentiated Gfi1^−/− PNEC Proliferate Poorly, but Respond Normally to Naphthalene-Induced Lung Injury

We have determined previously that Gfi1 is expressed in embryonic and adult PNEC and that lack of Gfi1 expression resulted in reduced numbers of of CGRP and SYN expressing cells.^28,31 In contrast, the number of Ash1 expressing cells in Gfi1^−/− mice remained the same.^28,31 As Ash1 is critically required for PNEC development,³⁴ Gfi1 might be required for maturation of presumptive PNEC.^28,31 Interestingly, some PNEC were able to differentiate in the absence of Gfi1, but could only weakly express CGRP and SYN. To determine the relevance of the NE phenotype to the post-naphthalene reconstitution process, we determined CGRP and SYN expression in naphthalene injured airway epithelium of wt and Gfi1^−/− mice 3 or 5 days following naphthalene administration. These time points are likely to be most informative for assessing NE differentiation.^6,12 Immunohistochemical detection of CGRP and SYN in oil-administered mice was used as a control (Figure 1a). Following naphthalene injury the number of PNEC increased.^6,12 In our experiments, control mouse CGRP and SYN-positive cells within bronchioles increased in numbers, peaking 5 days following naphthalene administration (Figure 1a).^6,12 Similar immunohistochemical analysis of Gfi1^−/− mice show that the 2–3-fold increase in numbers of CGRP and SYN positive foci in wt vs Gfi1^−/− mice was not statistically different (compare Figures 1a and b). Thus, whereas few PNEC differentiate without Gfi1, those that do are capable of responding to naphthalene-induced reconstitution signals.

The increase in NE cells post-naphthalene exposure does not rescue the impact of Gfi1^−/− on PNEC development. As expected, immunohistochemical staining showed a significant decrease of CGRP staining in oil-administered Gfi1^−/− mice than in the oil-administered wt mice (P =0.012) (Figure 1c). A similar decrease was observed in SYN expression in oil-administered Gfi1^−/− mice than in the oil-administered wt mice. In agreement with previous observations of post-naphthalene reconstitution,^6,12 a significant increase in CGRP expressing NE cell foci within bronchiolar epithelium was observed in wt mice 3 (P =0.016) and 5 (P =0.032) days following naphthalene administration compared with oil-treated animals (data not shown). A similar increase in CGRP-containing foci was seen in Gfi1^−/− mice 3 (P =0.002) and 5 (P =0.019) days following naphthalene administration compared with oil-administered animals (data not shown). However, the post-naphthalene environment did not rescue Gfi1^−/− defects, because when Gfi1^−/− mice were compared with wt mice 5 days following naphthalene administration, there was still a significant reduction in CGRP expressing NE foci (P =0.014) in the Gfi1^−/− animals (Figure 1c).

Similar to the increase in CGRP-positive NE foci following naphthalene administration, an increase in the wt SYN containing foci within bronchioli was observed compared with oil-treated mice. The increase in wt SYN-positive foci was significant 5 days following naphthalene administration compared with oil-treated mice (P =0.049) (data not shown). Gfi1^−/− mice also showed a significant increase in SYN-positive cells 5 days after naphthalene treatment when compared with oil-treated Gfi1^−/− animals (P =0.005) (data not shown). As expected from the differences in PNEC induced by Gfi1 loss, comparison of SYN staining in Gfi1^−/− mice and wt mice 5 days following naphthalene administration showed a significant difference between the two; Gfi1^−/− mice had less SYN-positive foci compared with wt mice (P =0.009) (data not shown).

A two-way ANOVA analysis indicated that whereas the increase in the number of CGRP or SYN expressing foci in each group (wt or Gfi1^−/−) is significantly different after naphthalene administration, there is no evidence that either wt or Gfi1^−/− PNEC respond differently to naphthalene administration. Thus PNEC, which have managed to differentiate without Gfi1 respond to naphthalene treatment in a similar manner to normal PNEC.

To examine further the impact of Gfi1 loss on NE differentiation we analyzed BrdU incorporation within CGRP + cells (PNEC and NEBs). We first counted BrdU + NEBs. In wt mice approximately one out of five NEBs (18%) initially revealed one or more BrdU-labeled cells, increasing to two out of three NEBs (69%) 3 to 5 days post-naphthalene (data not shown). Next, we evaluated each individual CGRP + cells for BrdU staining. The post-naphthalene increase in BrdU + CGRP + cells was significant (P =0.001), but transient (Figure 1d). In contrast, BrdU + NEBs were rarely seen in Gfi1^−/− mice, and after exposure to naphthalene only one out of three (35%) NEBs were also BrdU + (data not shown). Similar to wt littermates, the post-naphthalene increase in individual Gfi1^−/− BrdU + CGRP + cells was transient (Figure 1d). Notably, the 3-day post-napthalene BrdU staining in Gfi1^−/− airways was highly variable, and prevented the increase in BrdU + NE cells from reaching statistical significance. Yet, the number of BrdU + CGRP + cells was lower in Gfi1 KO mice than wt littermates in both oil-treated controls, and at either 3 or 5 days after naphthalene (P =0.05) (Figure 1d). The decreased BrdU labeling in Gfi1^−/− CGRP + cells is consistent with a role for Gfi1 in mediating proliferation of NE cells. However, the Gfi1^−/− impairment of NE cell proliferation was not impacted by naphthalene treatment, as the naphthalene-induced rate of increase of the BrdU + CGRP + cells was not statistically different between wt and Gfi1^−/− littermates (P =0.8). In other words, Gfi1^−/− airway epithelial cells that are capable of NE differentiation proliferate less well than Gfi1 expressing wt NE cells, but the naphthalene-induced rate of increase in the number of such cells was not impacted by loss of Gfi1.

Gfi1 Loss Decreases Post-Naphthalene Proliferation in Airway Epithelium

BrdU incorporation was also utilized to assess the role of Gfi1 loss on proliferation in the airway epithelium 3 and 5 days following naphthalene injury (Figure 2). As expected, there was a significant increase in the overall epithelial cell proliferation rate in wt animals (P<0.001) whereas the response in Gfi1^−/− mice was attenuated (P =0.058; 27 fold vs sevenfold, respectively) (Figure 2e). To characterize further post-naphthalene cellular dynamics, we focused on the terminal bronchioli, because this region comprises a putative regenerating zone harboring a stem cell niche.^18,20 Immunohistochemical analysis of oil-administered wt or Gfi1^−/− mice revealed few cells actively proliferating in terminal bronchioli (Figures 2a and b). Three days following naphthalene administration fewer cells were BrdU-positive in Gfi1^−/− mice compared with wt mice (Figure 2c and d). Notably, the observed induction of post-naphthalene proliferation in terminal bronchioli was not significantly different between Gfi1^−/− and wt littermates. Similar to Gfi1 effects on NE cell proliferation, the Gfi1^−/− decrease in BrdU staining in terminal bronchioli was not reversed by naphthalene. Thus, whereas Gfi1 deletion decreases airway epithelial proliferation, the significance of Gfi1 deletion upon post-naphthalene proliferation is restricted to the proximal and distal airways.

We next examined apoptosis in terminal bronchioli. After naphthalene administration, there was a statistically significant increase in apoptotic (TUNEL +) cells both in wt and Gfi1^−/− mouse airway epithelial cells (P<0.02 and P<0.05, respectively; Welch T-test) (Figure 2f). However, the difference between Gfi1^−/− and wt littermates in the post-naphthalene rate of increase in apoptosis was not statistically significant (P<0.1; Welch T-test) (Figure 2f). In conclusion, our analyses have uncovered Gfi1^−/− defects in airway epithelia and NE cell biology, but ANOVA statistical analyses indicated that these defects do not interact with naphthalene treatment.

Clara Cell Lung Reconstitution is Efficient in Gfi1^−/− Mice

Given a lack of statistical correlation between Gfi1^−/− defects and naphthalene treatment, we next examined whether lack of Gfi1 expression in murine lungs has an effect on Clara cell reconstitution. First, CC10 immunohistochemical staining of the distal airways 3 days following oil administration showed no initial difference between Gfi1^−/− and wt mice (Figures 3a and b). Moreover, Clara cells in the Gfi1^−/− airway are ablated similarly to wt littermate airway cells (Figure 3c). Finally, similar to wt mice, Clara cells in Gfi1^−/− lungs were fully reconstituted by 5 days following naphthalene administration (Figure 3d).

Quantification of CC10-positive cells showed no difference in staining between wt and Gfi1^−/− mice following oil treatment. A significant decrease was observed in both Gfi1^−/− and wt mice 3 days following naphthalene administration compared with oil-treated animals (P<0.001) (Figure 3e), confirming successful ablation with naphthalene. Wt and Gfi1^−/− mice showed very similar Clara cell reconstitution with no difference observed between the two, 5 days following naphthalene treatment (Figure 3e).