Main

BPD is a chronic lung disease that affects premature infants who have been treated with oxygen and mechanical ventilation(1). BPD has become an increasingly important clinical problem in neonatology as the numbers of small premature infants surviving the first few days of life has continued to increase. The pathogenesis of BPD is believed to be multifactorial with prematurity, barotrauma, and pulmonary oxygen toxicity having important roles. Pulmonary oxygen toxicity is a result of an imbalance between the concentration of reactive oxygen species and the antioxidant defense systems of the cell(2). Therefore an investigation of antioxidant enzyme gene expression in BPD and the level at which antioxidant enzyme gene expression is regulated in the premature infant is critical in understanding the molecular mechanisms that play an important role in pulmonary oxygen toxicity as a contributing factor to the development of BPD. To study antioxidant enzyme expression in BPD we used a premature baboon model; MnSOD was chosen as the enzyme to examine because in rodent models MnSOD has been shown to be critical in developing tolerance to the damaging effects of oxygen toxicity(3, 4).

The premature baboon model of BPD is a homolog for human infants with BPD. Clinical and morphologic studies have shown that the pathologic changes in the premature baboon model closely resemble the acute lung injury seen in human disease(5, 6). In addition to baboons with BPD, we investigated MnSOD expression in baboons in which a bacterial infection was superimposed on BPD. The baboon model of BPD plus infection results in more extensive alveolar disease than BPD alone(7, 8) and is relevant to human disease in which lung infections complicate the clinical course of BPD(9).

The BPD experimental group consisted of premature baboons delivered at 140± 2 d of gestation that were ventilated at high Fio2 for 16 d. In the BPD-infected experimental group, a bacterial infection was superimposed on hyperoxia-induced BPD by intratracheal instillation of 108Escherichia coli on d 11 of the 16-d ventilation. The following groups were used as controls: 1) 140-d gestation nonventilated animals (zero time control), 2) 156-d gestation nonventilated animals (gestational control), and 3) 140-d gestation animals ventilated for a 16-d period with clinically appropriate oxygen (PRN control).

We determined lung MnSOD mRNA and protein levels in this model system and found that the premature baboon lung is capable of mounting an antioxidant response and that the increase in MnSOD protein in BPD and BPD-infected animals is regulated, at least in part, at a posttranscriptional level.

METHODS

Animal studies. All animal care procedures were done according to the National Research Council's Guide for the Care and Use of Laboratory Animals and were approved by the animal care committees of the Southwest Foundation for Biomedical Research in San Antonio and the University of Texas Health Science Center at San Antonio. Details of the NICU animal care and the protocols for the animal groups used in this study were described previously(7, 8). To summarize briefly, baboons were delivered by hysterotomy at 140 or 156 ± 2 d of gestation (term 184 d). Each animal was assigned to a treatment group by random numbers. The zero time control group was killed at delivery at 140 ± 2 d of gestation. The gestational control group for the animals that would be treated for 16 d was delivered and immediately killed at 156 ± 2 d of gestation. The ventilatory controls for BPD and BPD-infection groups were delivered at 140± 2 d of gestation and immediately placed on positive pressure ventilation; this control group is designated PRN to indicate that these animals were given appropriate ventilatory and oxygen support to survive the premature delivery and birth. BPD premature baboons were delivered at 140± 2 d of gestation and mechanically ventilated for 16 d (Fio2 1.0 for 11 d, then O2 to keep Po2 at 40-50 torr for d 11-16). BPD-infected baboons were delivered at 140 ± 2 d of gestation and were placed on mechanical ventilation for 11 d at Fio2 of 1.0. At d 11 they had 108 E. coli intratracheally instilled and were maintained on O2 concentrations to keep Po2 at 40-50 torr for an additional 5 d. The lung pathology and morphology in the BPD baboon model has been previously described and documents the similarity between the lesions in these animals and those seen in neonatal humans with BPD(5, 6). The lesions are more severe in the animals where infection is superimposed on BPD(7, 8).

Dot blot quantitation of MnSOD RNA. To quantitate the level of MnSOD RNA we used densitometric analysis of a dot blot assay controlling for variations in loading and blotting by expressing the data relative to 18S rRNA. Total RNA was isolated from samples of baboon lung that had been stored at -70°C; RNA was extracted using TRI Reagent (MRC) in the protocol outlined by Molecular Research Center, Cincinnati, OH; this procedure is based on the single-step method of total RNA isolation(10). Total RNA was quantitated using absorbance at 260 nm and an extinction coefficient of 0.025 (μg/mL)-1cm-1. RNA application to nitrocellulose and hybridization were performed as described in the molecular cloning manual of Sambrook et al.(11). By applying increasing amounts of RNA to the filter we verified that we were working in a linear range of quantitation by scanning laser densitometry using ImageQuant software (Molecular Dynamics, Sunnyvale, CA). The data are given as relative densitometry units of MnSOD RNA per 18S rRNA. The data are averages of duplicate blots for each animal, and there were three to four animals in each group.

Northern analysis. Total RNA from the premature baboon lung was isolated as described above using TRI Reagent (MRC). The RNA was fractionated by electrophoresis in a 1% agarose-0.66 M formaldehyde gel and transferred to nitrocellulose. RNA ladders of 0.24 to 9.5 kb and 0.16 to 1.77 kb RNA (Life Technologies, Inc., Bethesda, MD) were used as a size markers. Hybridization and radioautography were performed as described previously(12).

Preparation of radioactively labeled probes for MnSOD RNA and 18S rRNA. To obtain a probe for MnSOD RNA, we used a human MnSOD cDNA in a pGem4 vector that was generously provided by Dr. David Goeddel of Genentech(South San Francisco, CA). A restriction endonuclease digest of pGem4-MnSOD using EcoRI and BamHI yielded a 750-bp cDNA fragment that was labeled with [32P]dCTP using random primers DNA labeling(13). This fragment corresponds to 66 bases of 5′-UTR, the entire 666-base coding region, and 18 bases of 3′-UTR. For use as a control on the dot blots 18S rRNA cRNA was labeled using Promega's transcription protocol for synthesis of high specific activity RNA with T7 polymerase and [32P]CTP (Promega, Madison, WI). The template for the transcription reaction was pT7 RNA 18S (Ambion, Austin, TX).

Quantitation of MnSOD and catalase protein. To quantitate the level of MnSOD and catalase protein, we used Western analysis followed by densitometric analysis of the HRP-stained band. Densitometry was performed on a Molecular Dynamics laser densitometer, and the data were quantitated using Image-Quant software. The data were expressed as densitometric units relative to the amount of total protein subjected to SDS-PAGE.

Baboon lung protein extracts were prepared by homogenization in 25 mM Tris-HCl, pH 7.0, buffer containing 0.1 mM EDTA, 1% Triton X-100, 40 mM KCl, 0.1 mM phenylmethylsulfonyl fluoride, 0.2 U/mL aprotinin, and 10 μg/mL leupeptin. Lung homogenate was centrifuged at 27,000 × g for 15 min. The protein concentration in the supernatant material was measured using Coomassie Plus assay reagent from Pierce (Rockford, IL) with BSA as a standard. Protein samples were subjected to electrophoresis on 8% SDS-PAGE gels in the Mini-Protean II cell and transferred to nitrocellulose in the Mini Trans-Blot cell (Bio-Rad, Hercules, CA). After blocking with gelatin and reacting with rabbit primary antiserum, the blot was assayed with goat anti-rabbit IgG conjugated to HRP. The HRP band was detected by purple color development with 4-chloro-1-naphthol and hydrogen peroxide.

The primary antibody used to assay MnSOD protein was rabbit anti-rat MnSOD serum. Rat liver MnSOD was purified as previously described(14) and was the antigen provided to Hazelton Labs(Vienna, VA) for antibody production. The primary antibody used to detect catalase protein was rabbit anti-catalase serum generated against bovine liver catalase.

RNA-protein binding assay. A gel retention RNA-binding assay was performed as previously described(15) with the modification that baboon lung homogenate was subjected to centrifugation at 100,000 × g, and the resultant supernatant material was used to assay RNA-binding activity. The 32P-labeled cRNA probe, RMS-NP, used for these studies was generated by in vitro transcription and corresponds to a 216-base cis element located in the 3′-UTR, 111 bases downstream of the rat MnSOD RNA stop codon.

Statistical analysis. For each parameter calculated from measurements, the values for individual animals were averaged per experimental group, and the SE of the group mean was calculated. The significance of difference between groups was determined by an unpaired t test(16) or by an analysis of variance(17, 18).

RESULTS

Analysis of baboon lung MnSOD mRNA. To verify that the human cDNA would cross-react with baboon MnSOD RNA, we performed Northern analysis. The 750-base MnSOD cDNA fragment used in these studies corresponds to 66 bases of 5′-UTR, the entire 666-base coding region and 18 bases of 3′-UTR. We found two major baboon RNA species hybridized to the human32 P-labeled MnSOD cDNA; the mobility of the bands indicates the baboon MnSOD RNAs are approximately 1 and 4 kb in length (Fig. 1). These sizes correspond to the size of the two major MnSOD RNAs in human tissue(19, 20).

Figure 1
figure 1

Northern blot. Total RNA from baboon lung was separated on a 1% agarose-0.66 M formaldehyde gel, blotted onto nitrocellulose, and hybridized with random primer 32P-labeled human MnSOD cDNA. The numbers(kb) on the right of the radioautograph indicate the size of the baboon lung MnSOD RNA species calculated relative to an RNA standard ladder placed on the same gel.

Full size image

A representative dot blot RNA analysis using the 32P-labeled human MnSOD cDNA probe is shown in Figure 2, and the results of densitometric quantitation are given in Table 1. There is no change in the content of MnSOD RNA from 140 to 156 d of gestation. This time interval represents 76 to 85% of gestation (term = 184 d). When the 140-d gestation baboons were delivered by hysterotomy and treated with appropriate ventilatory and oxygen support (PRN controls), there was a 2-fold increase in MnSOD mRNA. This result indicates that the premature baboon lung is capable of modulating the level of MnSOD RNA after birth. In the lungs of animals subjected to the hyperoxic protocol that causes BPD and in lungs of BPD animals superimposed with E. coli infection there is significantly less MnSOD RNA compared with the PRN controls. There was no significant difference between the 156-d gestational control and BPD or BPD plus infection indicating that, although the premature baboon is capable of up-regulating the level of lung MnSOD RNA in the postnatal period, hyperoxia impairs this response.

Figure 2
figure 2

RNA dot blot. Total lung RNA from baboon was dot blotted onto nitrocellulose as described in “Methods.” A representative radioautograph of the dot blot is shown. The experimental group is indicated at the right of each row, and the amount of RNA placed onto the nitrocellulose is indicated at the top of each column.

Full size image
Table 1 MnSOD mRNA content
Full size table

Analysis of premature baboon lung MnSOD protein. We verified that the anti-rat antibody would cross-react with baboon MnSOD protein by Ochterlony immunodiffusion analysis; the formation of precipitation lines indicated there was a reaction of identity between baboon protein and rat liver MnSOD; preimmune serum served as a control (data not shown). The western blot confirmed that rabbit anti-rat MnSOD antiserum reacts with pure rat liver MnSOD and the material in baboon lung homogenate that has the same electrophoretic mobility resulting in a HRP detectable band at approximately 22,000-D subunit molecular mass. Figure 3 is representative of a Western blot used to quantitate the abundance of MnSOD protein in premature baboon lung.

Figure 3
figure 3

Western blot of MnSOD. Baboon lung extract was subjected to SDS-PAGE and Western analysis as described in“Methods.” A representative HRP-stained Western blot is shown. The animal groups are indicated at the top of the gel. The arrow indicates the mobility of immunoreactive MnSOD protein with an apparent molecular mass of 22 kD.

Full size image

Table 2 shows the quantitative results of Western analysis. There was very little immunoreactive lung MnSOD protein at 140 or 156 d of gestation. The level of MnSOD protein did not change during this time interval in utero. Postnatally there was a 5-fold increase in MnSOD protein in the PRN controls compared with gestational controls. The level of MnSOD protein was also elevated in BPD animals compared with gestational and PRN controls. The greatest induction of MnSOD protein was observed in the animal model of BPD superimposed with infection; MnSOD immunoreactive protein was significantly increased in BPD-infected animals compared with all other groups.

Table 2 MnSOD protein content
Full size table

Analysis of baboon lung catalase protein. Western blot analysis showed that rabbit anti-catalase serum reacted with baboon lung catalase resulting in a HRP-dectecable band at approximately 69 000 D(Fig. 4). This band has the same electrophoretic mobility as pure catalase (data not shown) and is in agreement with the known subunit molecular weight for catalase. Figure 4 is representative of a Western blot used to quantitate the abundance of catalase protein in premature baboon lung. The results of quantitative densitometry are given inTable 3. In contrast to the level of MnSOD protein, there was appreciable catalase protein present at 140 and 156 d of gestation, and the amount of catalase was not different between the gestational and PRN controls. As found with MnSOD protein, lung catalase protein was induced in hyperoxia-exposed BPD and BPD-infected animals. The level of catalase immunoreactive protein was significantly greater in lung extract from BPD baboons superimposed with E. coli infection compared with all other groups; in addition, there was a qualitative change in the appearance of a catalase doublet when the Western blot was performed with lung extract from BPD-infected animals (Fig. 4, lane 5). The reason for this change is not known.

Figure 4
figure 4

Western blot of catalase. A HRP-stained immunoblot was obtained as described in “Methods.” The arrow indicates the mobility of immunoreactive catalase protein with an apparent molecular mass of 69 kD. The animal groups are indicated at the top of the gel.

Full size image
Table 3 Catalase protein content
Full size table

MnSOD RNA-binding protein activity. Baboon lung extract contains protein that binds to a 216-base cis element in the 3′-UTR of rat MnSOD RNA, and the binding activity is diminished in extracts from BPD or BPD-infected animals (Fig. 5A). MnSOD RNA-binding protein in baboon lung exhibited the same characteristics as we found in rat lung(15), namely a requirement for free sulfhydryl groups on the protein and specificity versus catalase RNA-binding protein. MnSOD RNA-binding protein in baboon lung is redox-sensitive; alkylation of sulfhydryl groups with N-ethylmaleimide or oxidation of free sulfhydryl groups with diamide inhibited the formation of MnSOD RNA-protein complexes (Fig. 5B). Competition studies showed the baboon lung MnSOD RNA-binding protein is specific; catalase RNA did not compete with MnSOD RNA for binding to the protein (Fig. 5C).

Figure 5
figure 5

MnSOD RNA-binding activity. In each panel, thearrow indicates the location of the RNA-protein binding complex and the animal group or assay condition is indicated at the top of the gel.(A) Lane 1 of the autoradiograph shows the formation of a complex between 32P-labeled MnSOD RNA probe, RMS-NP, and a 20-μg lung extract from a 156-d gestation baboon. Complex formation was diminished in lung extract from BPD or BPD-infected animals, lanes 2 and 3, respectively. The mobility of free probe is indicated at the bottom of the gel.(B) Lane 1 shows the mobility of the complex formed between MnSOD RNA and baboon lung protein. Treatment of the lung extract with 100 mM diamide(lane 2) or 10 mM N-ethylmaleimide (lane 3) completely blocked complex formation, indicating that the presence of free sulfhydryl groups on the protein is required for binding to MnSOD RNA. (C) The specificity of the gel shift assay was shown by competition studies in which incubation with unlabeled MnSOD RNA abolished radioautographic evidence of the complex formation (lane 2), but incubation with ≈ 100-fold molar excess of unlabeled catalase RNA, CAT-XP2(29), did not inhibit the interaction between baboon lung protein and 32P-labeled MnSOD RNA.

Full size image

DISCUSSION

The content of MnSOD protein in the premature baboon lung is very lowin utero but is induced 5-fold in the postnatal period in baboons that received appropriate ventilatory support (PRN). MnSOD protein is increased even further in the BPD and BPD-infected animals. These results are important because they indicate that, although the level of MnSOD protein is relatively low in the premature lung, the animals were able to mount an antioxidant response to a hyperoxic challenge. This finding is in agreement with studies in the preterm rat and guinea pig in which antioxidant enzyme activity increased in response to hyperoxic challenge(21–23). There appear to species differences, however, because preterm rabbits were unable to increase antioxidant enzymes during hyperoxic exposure(24). Neonatal rats in the early postnatal period also demonstrate a hyperoxia-induced increase in MnSOD enzyme activity shown to be regulated at the pretranslational level of RNA stability(12). This mechanism of regulation is different from the response of the premature baboon lung reported in this study where hyperoxia-induced MnSOD expression appears to be regulated at the level of protein, not RNA. Our measurements of MnSOD mRNA indicate that the premature baboon lung is capable of inducing MnSOD mRNA in PRN animals but that hyperoxic exposure in the model of BPD inhibits this response. Changes in the rate of RNA synthesis (transcription) or RNA degradation (stability) may account for the lack of increase in MnSOD RNA in the lungs from hyperoxia-exposed baboons; an examination of these levels of regulation is beyond the scope of this report.

To examine the specificity of the MnSOD response, we measured protein expression of a different antioxidant enzyme, catalase. Unlike MnSOD, there was appreciable catalase protein at 140 and 156 d of gestation, and the level was not different from the amount in postnatal PRN control lungs. These data suggest catalase protein is present earlier in gestation than MnSOD. As with MnSOD, the level of catalase protein was increased in lung extracts from both the BPD and BPD-infected baboons, providing another example of the capability of the premature baboon to mount an antioxidant response to hyperoxic challenge.

A comparison of MnSOD with surfactant expression is of particular relevance because studies in other species show that surfactant and antioxidant enzyme expression increase in a parallel fashion in the last 15% of gestation(25). King et al.(8) examined SP-A expression in lungs from the same baboons used in the experiments reported in this study. They found that SP-A RNA was significantly reduced in BPD and BPD-infected animals compared with PRN controls. This is analogous to our findings on MnSOD RNA and lends credence to the notion that SP-A and MnSOD RNA levels may be subject to the same control mechanisms. However, in contrast to our finding of elevated MnSOD and catalase protein levels in BPD and BPD-infected animals, lavage SP-A protein decreased in both of these groups, and total SP-A (lung tissue + lavage) decreased in the BPD-infected animals compared with PRN controls. These results indicate the hyperoxia-induced increase in antioxidant enzyme protein expression is, at least in part, specific.

Although transcriptional events may play a role in the regulation of the MnSOD RNA level in the baboon model of BPD, the divergence of response between lung MnSOD RNA (decreased) and protein (increased) in the BPD and BPD-infected animals indicates that, to a large degree, increased MnSOD gene expression in BPD and BPD-infected lungs is regulated at a posttranscriptional level. Translational control of gene expression can be defined as a change in the efficiency of translation, that is the number of amino acids polymerized per unit time per mRNA molecule(26). If we assume the level of MnSOD protein is determined primarily by rates of protein synthesis with negligible changes in rates of protein degradation, we can get a reasonable idea of MnSOD translational efficiency in the premature baboon lung by dividing the average protein content by average RNA content. The result of this calculation is shown in Table 4. It appears that there is elevated translational efficiency in all three postnatal groups with the biggest increase seen in the BPD and BPD-infected group. Experiments presented in this report cannot define the mechanism responsible for the increased MnSOD translational efficiency in BPD and BPD-infected baboon lungs. The increases could be caused by changes in protein synthesis initiation rate, elongation rate, changes in the distribution of mRNA between ribonucleoprotein particles and polysomes, or a combination of these factors. An intriguing possibility is that a trans-acting MnSOD-RNA binding protein plays a role in the regulation of translation. We used a previously described assay for MnSOD RNA-binding protein(15) and found an inverse relationship between MnSOD RNA-binding activity and the level of MnSOD protein. MnSOD protein was highest in the lung extracts from BPD and BPD-infected animals, but MnSOD RNA-binding activity was diminished in the lung extracts from BPD, and BPD-infected baboons compared with 156-d gestation controls. Although the physiologic function of the MnSOD RNA-binding protein is unknown, our observations are consistent with the notion that the MnSOD RNA-binding protein may be a trans-acting negative regulator of translation in the premature baboon lung. There is precedence in other systems for the paradigm of translational regulation by specific protein-RNA interactions(27).

Table 4 Relative MnSOD translational efficiency
Full size table

Although MnSOD and catalase protein are significantly higher in the BPD and BPD-infected animals, these animals were not protected against acute lung injury. It is possible that not all of the antioxidant enzyme protein represents enzymatically active protein. We cannot test this notion because baboon lung tissue from this study has been used and there is no additional material available. As Fridovich and Freeman(28) have pointed out, the team of antioxidant defense enzymes is mutually supportive, and perhaps other critical enzymes and nonenzymatic antioxidants are not up-regulated in the early premature lung. However, the most probable explanation for the acute lung injury lies in the fact that oxygen toxicity is but one part of the complex lung disease seen in BPD. Further study of the antioxidant status (enzymatic and nonenzymatic) as well as the relationship between antioxidants, surfactant, and lung growth are needed to get a better understanding of the molecular basis and potential regulation of BPD.