Abstract
We have investigated the relative efficacy of a range of natural and synthetic retinoids on the induction of alveolar regeneration in a dexamethasone-treated mouse model. The aim was to explore the roles of the different retinoic acid receptors using receptor-selective agonists and to determine whether other natural retinoids in addition to all-trans-retinoic acid (tRA) were effective. Dexamethasone treatment of newborn pups led to a reduced lung surface area and increased mean chord length. Subsequently, tRA induced alveolar repair, improved mean chord length, and improved the lung surface area to volume ratio. We found that 4-oxo-RA and a retinoic acid receptor (RAR) α-selective compound were as effective as tRA at inducing alveolar regeneration, with neither showing a significantly better efficacy. An RARβ-selective compound was also effective, whereas a RARγ-selective compound was not. Other retinoids, such as 9-cis-RA, 13-cis-RA, retinol, and a pan retinoid X receptor (RXR) agonist, do not induce significant responses. Neither did granulocyte colony-stimulating factor. We also showed that an RARβ-null mutant mouse line responded to dexamethasone by failing to develop alveoli appropriately and that tRA induced alveolar regeneration, suggesting that RARβ was not required for the regenerative response.
The discovery of the induction of alveolar regeneration by retinoic acid (RA) represents a major step forward in uncovering a potential therapy for diseases involving the loss of gas-exchanging surface area in humans (e.g., emphysema). This observation was made in the elastase-treated rat, which, due to the destruction of alveoli, has a vastly reduced gas-exchanging surface area. The subsequent administration of all-trans-RA (tRA) for 12 d resulted in the restoration of mean alveolar diameter (Lm) and surface area per lung volume (1).
Similar results using RA have been obtained in this and in other model systems. Belloni and colleagues (2) performed a similar experiment and obtained a 50% reversal of the elastase damage using tRA and a 70% reversal of the damage with its isomer, 9-cis-RA, but in a repeat of that experiment only a mild improvement in lung volume was obtained without any improvement in lung mechanics (3). The elastase-treated mouse has also been used, and tRA gave a 44% improvement in Lm (4).
Dexamethasone treatment of newborn rat and mouse pups has been used to generate emphysematous-like morphology of the lungs with vastly reduced gas-exchanging surface areas. Although this is a disrupted development model rather than a destruction model, treatment of 4-wk-old rats with tRA induced a partial recovery (5), and in 6-wk-old mice a full recovery was obtained, when surface area per 100 g body weight was the measured parameter (6). In the tight-skin mouse, a genetic model of emphysema that has enlarged air spaces with thinned or broken alveolar walls, tRA reduced alveolar size and increased alveolar number (5). After removal of the left lung in rats, the right lung responds by proliferating and expanding in volume and weight, and tRA administration improved these parameters (7). tRA can also protect the lungs of rats from O2 damage when tRA and O2 are administered simultaneously (8).
There have been studies in which tRA has had no effect. These studies have included dexamethasone-treated mice (9), elastase-treated mice (10, 11), elastase-treated rabbits (12), elastase-treated rats (13), bleomycin-treated rats (14), TNF-α–overexpressing mice (10), and smoking guinea pigs (15). It is likely that different model systems may not all respond to tRA and that differences in parameters such as dosing regimes, the time at which the animals were examined, and the method of dosing may help to explain these negative results.
RA is not the only compound that seems to be able to induce alveolar regeneration. Granulocyte colony-stimulating factor (GCSF) induced a 44% reduction in Lm in the elastase-treated mouse model, which was the same level of effect as tRA (4). Together these compounds showed synergy, and an even further reduction was obtained. In elastase-treated rats, viral transfection with hepatocyte growth factor (HGF) induced alveolar regeneration by increasing the number of proliferating cells and decreasing the number of apoptotic cells (16). Pulmonary function was improved in these HGF-treated animals. It would be of great interest if there was a link in developmental pathways between RA, GCSF, and HGF.
RA is the biologically active derivative of retinol (vitamin A). RA acts in the nucleus to induce changes in gene expression. It is synthesized in the cytoplasm of cells from its precursor retinol, which is distributed throughout the body via the bloodstream. Two classes of enzymes perform this synthetic reaction: the alcohol dehydrogenases or the short-chain dehydrogenases convert retinol into retinal, and then retinal is oxidized to RA by the retinaldehyde dehydrogenases (17). RA enters the nucleus, perhaps bound to its cytoplasmic binding protein, where it interacts with two classes of ligand-activated nuclear transcription factors, the retinoic acid receptors (RARs) and retinoid X receptors (RXRs). There are three RAR genes (α, β, and γ) (18) and three RXRs (α, β, and γ) (19). The RARs and RXRs act as heterodimers and recognize consensus sequences known as retinoic acid response elements in the enhancer sequences of RA-responsive genes.
It is important to understand the molecular mechanism of induction of alveolar regeneration by RA and which RAR is involved in the process so that downstream targets can be identified for further potential therapeutic intervention. The particular RAR involved has not been identified until now, but, on the basis of developmental studies, certain conclusions have been made. All the components of the RA signaling pathway are present during alveolar development—the binding proteins, the RA synthesizing enzymes, the RARs and RXRs, and retinoids (20, 21). When null-mutant mice for the RARs were generated, the phenotypes were very revealing. The RARα null-mutant mouse had normal alveoli at postnatal Day 14, but by Day 50 alveolar volumes were higher, there were fewer alveoli, and the alveolar surface area was lower than normal (22). This suggests a role for RARα in the later, continuing process of alveologenesis, which progresses slowly until the age of 5–6 wk in rodents. The RARβ null-mutant mouse had smaller alveolar volumes, more alveoli, and a larger alveolar surface area postnatally and up to Day 64 (23). The RARγ null-mutant mouse had a decrease in alveolar wall volume density, alveolar surface area, and number of alveoli and a corresponding increase in Lm (24). These observations led to the idea that RARγ is the important positive regulator of alveolar septation, perhaps with a minor contribution from RARα in adulthood, and that RARβ is the important negative regulator. This does not necessarily mean that the same is true for the induction of regeneration, and in the work reported here we examine the roles of the individual RARs in regeneration.
We have used receptor-selective agonists to determine whether they have differential regenerative abilities. These compounds are synthetic retinoids with higher affinities for one or other of the RARs and have been widely used in retinoid studies to great effect. We have examined the efficacy of a range of retinoids, naturally occurring and synthetic, for their efficacy at inducing alveolar regeneration in a dexamethasone-treated mouse model. We found that 4-oxo-RA, tRA, and a RARα agonist consistently induce alveolar regeneration in various measures of regeneration, including decrease in Lm and increase in surface-to-volume ratio. Other retinoids tested were ineffective. We also show that GCSF does not induce alveolar regeneration in this system and that the RARβ null-mutant mouse responds to RA.
Pregnant mice of the outbred TO strain were obtained from Charles River. The RARβ null-mutant mice were bred from a colony obtained from Professor P. Chambon (Strasbourg, France). Male and female mice were used in group sizes varying from 6 to 21 animals. Postnatal Day 0 (P0) was defined as the day of birth.
Dexamethasone (Sigma, Poole, UK) was dissolved in PBS and given as a 10-μl subcutaneous injection at a concentration of 40 μg/ml. This was administered for 10 d (from P4 to P13). Control mice received PBS. On Day 35, various retinoids were administered for 10 d, with a 2-d break on P40 and P41. This break in dosing was an attempt to prevent a full induction of the CYP enzymes, which catabolize RA and would thus reduce the efficacy of the administered retinoids. The retinoids were dissolved at a concentration of 2 mg/ml in dimethylsulphoxide/peanut oil (1:1), and a dose of 2 mg/kg was administered by intraperitoneal injection. A 10-fold higher dose of all-trans-retinol was administered because this precursor of tRA is known to be far less potent. GCSF (PeproTech, London, UK) was administered, as in Ishizawa and colleagues (4), by subcutaneous injection at a concentration of 50 μg/kg/d. Control mice received DMSO:peanut oil only. tRA, 9-cis-retinoic acid, 13-cis-retinoic acid, and all-trans-retinol were obtained from Sigma; the RAR agonists were obtained from CIRD Galderma (Sophia Antipolis, France); and 4-oxo-RA and the RXR agonist were obtained from Hoffman-La Roche (Basel, Switzerland). Animals were kept for 4 wk after the end of the retinoid dosing (until Week 13) and killed.
The animals were weighed, and the lungs were carefully removed and intubated with a 22-gauge cannula. The lungs were inflated to a standard pressure of 20 cm H2O with 4% paraformaldehyde, after which the trachea was ligated and the tissue placed in 4% paraformaldehyde for 48 h. The lungs were washed in water and placed in 70% alcohol, and the lung volumes were measured by fluid displacement. The lungs were processed and embedded in paraffin wax, and 5-μm sections were stained with hematoxylin and eosin and examined by light microscopy. Sections were analyzed on a light microscope linked to a digital imaging system. Images were acquired and processed using Image ProPlus software (Data-Cell Ltd, Finchampstead, UK) to calculate the Lm. At least five nonoverlapping fields per section from six sections were analyzed for each animal. Following Weibel (25), knowledge of the Lm and lung volume allows gas-exchanging surface area (Sa) and the surface-to-volume ratio to be calculated.
Control mice that had been treated with PBS and then peanut oil/DMSO showed lung histology and Lm measurements that were indistinguishable from normal, uninjected mice (not shown), confirming that these vehicles had no effect. Over the course of these experiments, two different control groups were used that were of different ages. As a result, their weights, lung volumes, and Sa were different, but their Lm values and Sa/volume ratios were identical (control 1 and control 2 in Table 1). For this reason, Lm and Sa/volume ratio were chosen as comparative parameters, which eliminates variation due, for example, to weight changes. In the latter case, the two control groups had very different Sa/100 g body weight parameters (1,281 cm3/100 g and 953 cm3/100 g) caused by body weight differences, and this could confound the retinoid comparisons. Therefore, even though we had previously used this parameter in our studies (6), we did not use it here.
Retinoid | n | Average Weight (g) | Average Lm (μm) | % Recovered | Average Volume (cm3) | Average Sa (cm2) | Average Sa/Volume Ratio | % Recovered |
|---|---|---|---|---|---|---|---|---|
| Control 1 | 10 | 34.3 ± 1.3 | 68.1 ± 0.9 | 100 | 0.740 ± 0.015 | 434.7 ± 7.5 | 589.0 ± 7.8 | 100 |
| Dexamethasone | 21 | 32.7 ± 0.6 | 94.3 ± 2.1 | 0 | 0.717 ± 0.028 | 305.3 ± 10.5 | 428.5 ± 9.5 | 0 |
| tRA | 8 | 28.3 ± 1.7 | 81.6 ± 2.0 | 49 | 0.702 ± 0.038 | 345.4 ± 18.1 | 492.7 ± 13.5 | 41 |
| 9-cis-RA | 15 | 26.4 ± 1.1 | 89.1 ± 2.5 | 20 | 0.727 ± 0.059 | 320.1 ± 26.8 | 442.7 ± 17.2 | 9 |
| 13-cis-RA | 15 | 30.5 ± 0.5 | 98.7 ± 3.1 | 0 | 0.749 ± 0.040 | 308.4 ± 20.1 | 420.4 ± 13.8 | 0 |
| 4-oxo-RA | 6 | 25.8 ± 0.8 | 78.5 ± 1.7 | 60 | 0.700 ± 0.018 | 336.5 ± 17.5 | 482.8 ± 29.3 | 34 |
| tRol | 6 | 32.1 ± 2.4 | 91.0 ± 1.3 | 13 | 0.766 ± 0.038 | 337.7 ± 19.5 | 440.1 ± 6.4 | 8 |
| RARα | 15 | 25.1 ± 0.7 | 82.8 ± 2.8 | 44 | 0.633 ± 0.050 | 302.2 ± 18.5 | 490.7 ± 16.1 | 40 |
| RARβ | 15 | 29.2 ± 0.6 | 84.6 ± 2.3 | 37 | 0.579 ± 0.035 | 274.8 ± 13.8 | 479.9 ± 11.6 | 33 |
| RARγ | 7 | 24.69 ± 0.7 | 87.06 ± 2.3 | 28 | 0.734 ± 0.047 | 338.7 ± 24.0 | 465.8 ± 14.4 | 24 |
| RXR | 8 | 29.0 ± 1.5 | 94.0 ± 2.5 | 0 | 0.561 ± 0.048 | 242.6 ± 20.9 | 436.3 ± 13.7 | 5 |
| Control 2 | 10 | 42.7 ± 3.7 | 68.9 ± 3.9 | 100 | 0.690 ± 0.028 | 402.5 ± 19.7 | 581.8 ± 10.7 | 100 |
| GCSF | 6 | 28.66 ± 2.3 | 91.19 ± 2.27 | 12 | 0.733 ± 0.034 | 317.75 ± 15.7 | 433.2 ± 8.9 | 3 |
A total of 21 animals were treated with dexamethasone. Although dexamethasone had an effect on body weight immediately after dosing (6), these animals had almost caught up with the control animals by the time the experiment had terminated (Table 1). Dexamethasone had a dramatic effect on Lm, increasing the length from 68 to 94 μm (40%) (Figure 1, Table 1). Sa and the Sa/volume ratio was reduced by ∼ 30% (Table 1). A typical dexamethasone-treated lung is shown in Figure 4B; a typical control lung in is shown in Figure 4A.
Figure 1. Graph of Lm (in μm) in response to a range of different treatments. Above the shaded line is the range in which lungs show no response to the treatments, between the shaded and clear lines is the range in which lungs show a response to the treatment but have not fully regenerated their Lm values, and below the clear line is the range in which lungs respond fully to the treatment. Stars mark groups that are significantly different from dexamethasone-treated animals. Bars = SEM.
[More] [Minimize]A range of naturally occurring retinoids was administered to dexamethasone-treated animals, and their efficacies relative to tRA were determined by comparing Lm values. In addition, the number of animals regenerating and responding to the compounds in each group were compared because groups of animals did not all behave identically to these treatments.
A group of eight animals was treated with tRA. The average Lm value of this group was 82 μm, which represents a recovery (49%) from 94 μm in the dexamethasone-treated animals (Table 1, Figure 1). What this average value hides is that there was considerable variation in the response of the animals to tRA. Of the eight animals in this group, two animals regenerated completely as defined by a Lm value close to the Lm value obtained from the control mice (Lm, 69–75 μm), four animals responded partially (defined as a Lm of 75–85 μm), and two animals did not respond (defined as a Lm of > 85 μm). These zones are marked on the graph in Figure 1. Thus, in terms of the percent numbers regenerating, this gives tRA a value of 25% of the group showing complete regeneration or 75% of the group showing some response (see Figure 2). An example of a regenerated lung is shown in Figure 4C.
Figure 2. Graph of the responding animals in each treatment group (dark bars) and the fully regenerating animals (light bars) in each treatment group. Responding and regenerating animals are defined as in Figure 1 (i.e., lowering the Lm or returning it to the normal range, respectively).
[More] [Minimize]A group of 15 animals was treated with 9-cis-RA to determine how this compound compared with tRA. There was a small improvement in Lm (89 μm), which represents a 20% return toward control levels (Table 1, Figure 1), and statistically there was no difference between the dexamethasone and 9-cis-RA Lm values (P = 0.468). In this group there were two cases in which the Lm values returned to the regenerating range, representing 13% of animals showing complete regeneration (Figure 2) and two cases that were in the partially responding range, giving a total response of 26% of animals. Thus, by either of these criteria, 9-cis-RA is far less active than tRA. An example of a 9-cis-RA–treated lung is shown in Figure 4D.
13-cis-RA was tested as a comparison of RA isoforms in a group of 15 animals. There was no improvement in the Lm values of the 13-cis-RA group compared with the dexamethasone group (they were slightly worse) (Table 1, Figure 1). There was one animal that had a Lm value in the complete regeneration range, giving a percent regeneration of 7% (Figure 2). No other animal showed any response, so there was only a figure of 7% for the total number of animals responding. A typical 13-cis-RA–treated lung is shown in Figure 4E.
A group of six animals was treated with 4-oxo-RA. There was a clear reduction in average Lm compared with dexamethasone-treated levels, nearly back to the normal Lm value (from 94.3 to 78.5 μm) (Table 1, Figure 1). This represents a recovery of 60% back to control levels (Table 1). Not all the members of the group behaved in the same fashion: One animal showed complete regeneration by returning the Lm to a value in the 69–75 mm range, giving a value of 17% of cases showing complete regeneration (Figure 2), and four cases had Lm values in the partial response range, giving a responding value of 83% of the cases (Figure 2). An example of a 4-oxo-RA–induced regenerate is shown in Figure 4F.
Finally, a group of six animals was used for the administration of all-trans-retinol, the precursor of tRA. Because this is a far less potent compound than the acid derivatives in many biological assays, it was administered at 10× the concentration of the other retinoids. However, even this high dose had no effect on Lm (Figure 1): There was no significant difference between retinol treatment (Lm = 91.0 μm) and dexamethasone treatment (Lm = 94.3 μm) (P = 0.409). No individual animal showed any return of Lm value close to control levels, nor did any animal have an Lm value that entered the responding range. Thus, this group has a 0% value (Figure 2).
In the next experiment, the efficacies of four receptor-selective agonists were compared with the naturally occurring retinoids described previously. Fifteen animals were used for a group to which a RARα agonist was administered at the same concentration as tRA. On average, this treatment reduced the Lm to 83 μm, a 44% recovery (Table 1, Figure 1). This value is close to that for tRA, but it was not as effective as 4-oxo-RA. However, this figure hid the fact that individually there was a successful induction of regeneration in four cases, giving a frequency of complete regeneration of 27% of cases (Figure 2), and a further six cases that partially regenerated in response to the RARα, giving a responding value of 67% of cases (Figure 2). An example of a RARα induced regenerated lung is shown in Figure 4G.
Another group of 15 animals was administered a RARβ agonist. This treatment had a significant effect on Lm, reducing it to 85 μm, a value similar to that for the RARα compound and tRA (Table 1, Figure 1), and giving a recovery value of 37%. Included in this average value was one case where the Lm had fully regenerated, giving a figure of 7% of cases, showing complete regeneration (Figure 2), and nine cases that showed some partial regeneration in response to the agonist, giving a responding value of 67% of cases. An example of a RARβ agonist treated lung is shown in Figure 4H.
A group of seven animals was administered a RARγ agonist, which reduced the average Lm to 87 μm, a value that was worse than tRA, 4-oxo-RA, and the RARα and RARβ agonists (Table 1, Figure 1). This figure gives a value of 28% recovery (Table 1). Statistically, there was no significant difference between the Lm values of this group compared with the dexamethasone-treated group (P = 0.0685). None of the animals in this group were deemed to have regenerated, thus giving a rate of 0% in Figure 2, although four of the seven animals showed a partial response, giving a frequency of response of 57% of cases (Figure 2).
Finally, a group of eight animals was used for the administration of a pan-RXR agonist. This compound had no effect on the average Lm, which was identical to the dexamethasone-treated group (Figure 1, Table 1). None of these animals regenerated, giving a 0% regeneration rate (Figure 2), and only one animal showed a response, giving a responding value of 13% of cases (Figure 2). An example of a RXR agonist–treated lung is shown in Figure 4I.
In summary, on the basis of Lm measurements, the relative efficacy of these retinoids is 4-oxo-RA > tRA > RARα agonist > RARβ agonist, and the compounds that had no statistically significant effect on Lm values were RARγ agonist, 9-cis-RA, all-trans-retinol, 13-cis-RA, and the RXR agonist. However the RARγ agonist and 9-cis-RA induce some response in individual cases (Figure 2).
In Table 1, the other parameters measured on the lungs of the experimental animals are recorded, including body weights, lung volumes, surface area (Sa), and Sa/volume ratios. Considering the Sa values, none of the compounds returned the lungs to anywhere near the average control values of 434 cm3. The most successful was tRA, which gave an average value of 345 cm3. The RARβ− and the RXR-treated animals had smaller lung volumes than the dexamethasone-treated group.
The weights of the animals that had been treated with active retinoids were less than control animals or dexamethasone-treated animals (Table 1). A previous measure of regeneration that we had used was Sa/100 g body weight (6), but this is susceptible to variation in body weight due to ageing because the lung volume per body mass decreases as animals age. This can be seen in Table 1 in the two control groups. Control 2 group was older and heavier and had a 25% decrease in Sa/100 g body weight (data not shown). Using Sa/volume ratios, however, the two control groups had virtually identical values (589 and 581); for this reason, this parameter was used to compare groups in Table 1 and to calculate another value of % recovery. The Sa/volume ratios are drawn as a graph in Figure 3. From these data, the relative efficacies of the different retinoids were found to be tRA and RARα agonist > 4-oxo-RA and RARβ agonist. Although the RARγ agonist induced a 24% recovery in Sa/volume, it did not generate a statistically significant difference compared with the dexamethasone group (P = 0.0546), and the other compounds (9-cis-RA, tRol, RXR agonist, and 13-cis-RA) had no statistically significant effect (Figure 3).
Figure 3. Graph of the surface area to volume ratios in response to a range of different treatments. The stars mark the groups that are significantly different from dexamethasone-treated animals. Bars = SEM.
[More] [Minimize]In view of the recent data showing that GCSF is as effective as tRA at inducing recovery in Lm in the elastase-treated mouse model (4), it was of interest to examine whether this compound had an effect on the dexamethasone-treated mouse model. A group of six animals was treated with GCSF for 2 wk. There was no significant improvement in Lm (Table 1, Figure 1) or in the Sa/volume ratio (Table 1, Figure 3). One animal showed a partial regeneration response, giving some improvement in Lm, and therefore recorded a value of 0% regeneration and 17% of cases responding (Figure 2).
It is of interest to determine whether any of the retinoic acid receptor knockout lines respond to tRA by regenerating, assuming that they have developed alveoli in the first place. This has considerable relevance with regarding the downstream pathway of RA-induced regeneration involved in the response. We have only used the RARβ null-mutant line, which initially has more alveoli than normal (23). Lm measurements on 10 P40 knockout mice were performed and compared with the same number of normal TO mice. There was no difference between the two Lm values (TO = 68.1 μm, βko = 67.9 μm). As emphasized by Massaro and colleagues (23), Lm is not a sensitive measure, and in that work alveolar volumes were used instead to detect differences. When treated with dexamethasone (n = 6), these RARβ knockout animals had greatly increased Lm values (96.5 μm). tRA treatment of eight dexamethasone-treated animals resulted in a decrease in Lm value (86.3 μm), a 35% recovery (Figure 1). Within this group, two animals showed complete regeneration, giving a 25% of cases regeneration rate (Figure 2), and two animals showed partial regeneration giving a 50% of cases responding. The remaining four animals did not respond.
Typical examples of lungs after retinoid treatment are presented in Figure 4. One of the most striking observations on these lungs was the difference in wall thickness between the groups. A control lung is shown in Figure 4J at high power, showing normal-sized alveoli and their typical cellular structure. In Figure 4K, a typical dexamethasone-treated lung is shown, revealing a larger alveolar size and thin walls. This seems to be the most obvious structural result of dexamethasone treatment. Figure 4L shows a typical regenerated lung after treatment with an RARα agonist. The alveoli have decreased in diameter compared with Figure 4K, and the walls are highly cellular and much thicker. There seemed to be many capillary spaces in these thick walls, which gave them a bubbly appearance, although this impression has not been verified with endothelial cell markers.
Figure 4. Typical histologic images of lungs from various treatment groups. (A) Control PBS, DMSO/oil-treated lung. (B) Dexamethasone-treated lung. (C) tRA-treated lung. (D) 9-cis-RA–treated lung. (E) 13-cis-RA–treated lung. (F) 4-oxo-RA–treated lung. (G) RARα agonist–treated lung. (H) RARβ agonist–treated lung. (I) RXR agonist–treated lung. (J) High power of a control lung. (K) High power of dexamethasone-treated lung. (L) High power of a RARα agonist–treated lung.
[More] [Minimize]We have used a dexamethasone-treated mouse model to study the effects of a range of different retinoids, natural and synthetic, for their ability to induce regeneration of alveoli. Dexamethasone, when administered to rats during the first 2 wk after birth, permanently impairs alveolization (26, 27), and we have found the same dose-dependent phenomenon in the mouse (6). The net result of such a treatment is a lung with a greatly increased Lm value and strikingly thin walls (Figure 4K), as has been noted in studies reporting that dexamethasone treatment accelerates the thinning of the alveolar walls (28, 29). We have previously shown this model system to be responsive to tRA because new septation is induced, resulting in an increased number of alveoli, a decreased Lm value, and an increased gas exchanging surface area/100 g body weight (6), as has previously been seen in the elastase-treated adult rat (1).
Retinoids are a large family of molecules with varying biological potencies due to their varying abilities to activate RARs. Of the naturally occurring retinoids, such as tRA, 9-cis-RA, 13-cis-RA, and 4-oxo-RA, tRA is generally the most potent and activates each of the three RARs (RARα, RARβ, and RARγ). Because of this multiple receptor activation, the effect of RA on cells is pleiotropic (30). In vivo, RA has several therapeutic indications, but its administration has side effects, including lipid and bone toxicity, and it is highly teratogenic (31, 32). To counteract these unwanted effects, receptor-selective agonists have been developed (33–38) with the intention of activating only subsets of genes in the retinoid pathway. This they indeed do, with specific RAR agonists having effects on, for example, gene induction (39), promotion of growth arrest and cell death (40, 41), epithelial metaplasia (42), or neurite induction (43). They also show different teratogenic profiles in vivo (44–47), with RARγ agonists being the least teratogenic (48), and they show different morphogenic induction profiles in embryos (49). These agonists are therefore useful compounds for dissecting the roles of individual receptors in particular retinoid effects, and for this reason we examined the efficacy of a RARα, RARβ, and RARγ agonist compared with tRA in our mouse model system. We also examined several natural retinoids whose efficacy varies in various cellular systems to see if that was the case here in an in vivo system. Only tRA and 9-cis-RA have been used in animal studies.
Using Lm as a measure and calculating the recovery of this parameter, tRA gave an average of 49% recovery. 4-oxo-RA, a natural retinoid that is a product of the CYP26 enzymes, gave 60% recovery. The next most effective retinoid was a RARα agonist, which gave 44% recovery, followed by a RARβ agonist with 37% recovery. The order of efficacies was thus 4-oxo-RA > tRA > RARα agonist > RARβ agonist. The remaining compounds were not statistically significantly different from the dexamethasone-treated animals, even though the RARγ agonist gave a 28% recovery and 9-cis-RA gave 20% recovery. The compounds that had no effect on Lm were all-trans-retinol (the parent vitamin A molecule), a pan RXR agonist, and GCSF. The latter was used because of the recent report of its efficacy in an elastase-treated mouse model system (4), but in this dexamethasone model it seemed to have no effect on Lm values, suggesting that this phenomenon may be model dependent. We also showed in these Lm studies that the RARβ null-mutant strain of mice responded to dexamethasone in the same way as normal mice by increasing their Lm and decreasing the number of alveoli. When treated with tRA, this group of animals responded by showing a 35% recovery. We can therefore conclude that RARβ is not crucial to the induction of alveolar regeneration by tRA.
Using surface area/volume ratios as a measure, the relative efficacies we obtained were similar. The order of efficacies was tRA and RARα agonist > 4-oxo-RA and RARβ agonist. The other compounds did not have a statistically significant effect using this parameter, although some of them induced a degree of recovery. For example, the RARγ agonist gave a 24% recovery using this parameter.
The active natural retinoids were thus tRA and 4-oxo-RA. The latter is a breakdown product of tRA, generated by the CYP26 enzymes (50–52), which suggests that these enzymes may not be present simply to catabolize tRA but may make biologically active products. Indeed, 4-oxo-RA is active in other in vivo situations. It alters anteroposterior patterning in the Xenopus embryo through the activation of RARβ (53, 54), rescues defects produced in avian embryos by a lack of vitamin A (55, 56), alters epidermal thickness in mouse skin (57), and restores germ cell production in vitamin A–deficient mice (58). 4-oxo-RA also has many effects on various cell lines, and its effect on promyelocytic leukemia cells is thought to operate through the RARα receptor (59). We have found that the induction of the Cyp26A1 and B1 genes by tRA operates through the RARα receptor (56), so both of these sets of data fit well with the results of the receptor agonist studies reported here on the lung.
These agonist results revealed that the pan RXR agonist was inactive, whereas the RARα and RARβ agonists were effective. This implies that the mechanism of induction of alveolar regeneration by retinoids involves the activation of a RAR and not a RXR alone, which is a common finding in retinoid biology because a RAR/RXR heterodimer is the normal receptor configuration. To determine which is the active heterodimer in alveolar regeneration, we need to show synergy between a RAR agonist and a RXR agonist. Which RAR is the active component was not completely established by using the agonists because the RARα and the RARβ agonists were effective. However, we showed that RARβ is not crucial to the regenerative process because it could take place in the absence of this receptor (i.e., in the RARβ null-mutant mouse line), so we are led to the conclusion that RARα is the active RAR in alveolar regeneration.
A potential role for RARβ as a negative regulator of alveolar development (23) has been developed based on studies in the RARβ null-mutant mouse line and the administration of a RARβ agonist. After these treatments, more alveoli than normal developed, and it is possible therefore that the administration of a RARβ antagonist alone might induce alveolar regeneration. However, it was concluded by Massaro and colleagues (23) that RARβ is not an inhibitor of alveolus formation after the perinatal period. In accordance with this, we did not find any evidence to suggest that RARβ is a negative regulator of alveolar regeneration because the null-mutant mouse did not regenerate more successfully than normal. The fact that RARα has been shown to play a role in alveolus formation not in the immediate postnatal period but between the ages of 14 and 50 d (22) supports our agonist data because this is precisely the time that we began retinoid administration. We conclude that RARα is the receptor involved in the regenerative response.
It could be argued that the concentration of these agonists in these studies was too high, being the same as that used for tRA. One might have expected a considerably lower dose to be effective because of the higher efficacies of these selective agonists in most biological situations. Perhaps at these concentrations the individual receptors were saturated, and therefore the intended selectivity was not obtained. These experiments need to be repeated with decreasing doses of the agonists to confirm which is the most effective. Another problem that arose in these studies was the fact that a maximum of 25% of the animals in a group regenerated fully (tRA and RARα groups). However, a proportion of the remaining treated animals in the 4-oxo-RA group often responded to the retinoids and partially regenerated giving, at maximum, an 80% response rate (a figure that includes the fully regenerated cases). The ideal situation is for all the animals to regenerate completely, and we are currently investigating whether this is a function of dose, time after administration, method of administration, or individual variability in uptake. It could also be due to the fact that we used an outbred strain of mice for these experiments, and we are currently repeating these experiments using inbred strains. Perhaps variation in some of these parameters may be responsible for the existence of reports of the failure of tRA to induce regeneration, and if these discrepancies could be resolved it would be of enormous benefit to the subject of alveolar regeneration.
The author thanks Professor P. Chambon, Strasbourg, France for providing the RARβ null-mutant mice to found a colony; CIRD Galderma, Sophia Antipolis, France for CD compounds; and Hoffmann-La Roche, Basel, Switzerland for several RA isoforms.
| 1. | Massaro GD, Massaro D. Retinoic acid treatment abrogates elastase-induced pulmonary emphysema in rats. Nat Med 1997;3:675–677. |
| 2. | Belloni PN, Garvin L, Mao CP, Bailey-Healy I, Leaffer D. Effects of all-trans-retinoic acid in promoting alveolar repair. Chest 2000;117(Suppl 1)235S–241S. |
| 3. | Tepper J, Pfeiffer J, Aldrich M, Tumas D, Kern J, Hoffman E, McLennan G, Hyde D. Can retinoic acid ameliorate the physiologic and morphologic effects of elastase instillation in the rat? Chest 2000;117 (Suppl 1)242S–244S. |
| 4. | Ishizawa K, Kubo H, Yamada M, Kobayashi S, Numasaki M, Ueda S, Suzuki T, Sasaki H. Bone marrow-derived cells contribute to lung regeneration after elastase-induced pulmonary emphysema. FEBS Lett 2004;556:249–252. |
| 5. | Massaro GD, Massaro D. Retinoic acid treatment partially rescues failed septation in rats and in mice. Am J Physiol Lung Cell Mol Physiol 2000;278:L955–L960. |
| 6. | Hind M, Maden M. Retinoic acid induces alveolar regeneration in the adult mouse lung. Eur Respir J 2004;23:20–27. |
| 7. | Kaza AK, Kron IL, Kern JA, Long SM, Fiser SM, Nguyen RP, Tribble CG, Laubach VE. Retinoic acid enhances lung growth after pneumonectomy. Ann Thorac Surg 2001;71:1645–1650. |
| 8. | Veness-Meehan KA, Bottone FG Jr, Stiles AD. Effects of retinoic acid on airspace development and lung collagen in hyperoxia-exposed newborn rats. Pediatr Res 2000;48:434–444. |
| 9. | Srinivasan G, Bruce EN, Houtz PK, Bruce MC. Dexamethasone-induced changes in lung function are not prevented by concomitant treatment with retinoic acid. Am J Physiol Lung Cell Mol Physiol 2002;283:L275–L287. |
| 10. | Fujita M, Ye Q, Ouchi H, Nakashima N, Hamada N, Hagimoto N, Kuwano K, Mason RJ, Nakanishi Y. Retinoic acid fails to reverse emphysema in adult mouse models. Thorax 2004;59:224–230. |
| 11. | Lucey EC, Goldstein RH, Breuer R, Rexer BN, Ong DE, Snider GL. Retinoic acid does not affect alveolar septation in adult FVB mice with elastase-induced emphysema. Respiration (Herrlisheim) 2003;70:200–205. |
| 12. | Nishi Y, Boswell V, Ansari T, Piprawala F, Satchi S, Page CP. Elastase-induced changes in lung function: relationship to morphometry and effect of drugs. Pulm Pharmacol Ther 2003;16:221–229. |
| 13. | March TH, Cossey PY, Esparza DC, Dix KJ, McDonald JD, Bowen LE. Inhalation administration of all-trans-retinoic acid for treatment of elastase-induced pulmonary emphysema in Fischer 344 rats. Exp Lung Res 2004;30:383–404. |
| 14. | Segel MJ, Or R, Tzurel A, Lucey EC, Goldstein RH, Izbicki G, Breuer R. All-trans-retinoic acid (ATRA) is of no benefit in bleomycin-induced lung injury. Pulm Pharmacol Ther 2001;14:403–407. |
| 15. | Meshi B, Vitalis TZ, Ionescu D, Elliott WM, Liu C, Wang XD, Hayashi S, Hogg JC. Emphysematous lung destruction by cigarette smoke: the effects of latent adenoviral infection on the lung inflammatory response. Am J Respir Cell Mol Biol 2002;26:52–57. |
| 16. | Shigemura N, Sawa Y, Mizuno S, Ono M, Ohta M, Nakamura T, Kaneda Y, Matsuda H. Amelioration of pulmonary emphysema by in vivo gene transfection with hepatocyte growth factor in rats. Circulation 2005;111:1407–1414. |
| 17. | Duester G. Families of retinoid dehydrogenases regulating vitamin A function: production of visual pigment and retinoic acid. Eur J Biochem 2000;267:4315–4324. |
| 18. | Kastner P, Chambon P, Leid M. Role of nuclear retinoic acid receptors in the regulation of gene expression. In: Blomhoff R, editor. Vitamin A in health and disease. New York: Marcel Dekker Inc.; 1994. pp. 189–238. |
| 19. | Kliewer SA, Umesono K, Evans RM, Mangelsdorf DJ. The Retinoid X receptors: modulators of multiple hormonal signaling pathways. In: Blomhoff R, editor. Vitamin A in health and disease. New York: Marcel Dekker Inc.; 1994. pp. 239–255. |
| 20. | Hind M, Corcoran J, Maden M. Temporal/spatial expression of retinoid binding proteins and RAR isoforms in the postnatal lung. Am J Physiol Lung Cell Mol Physiol 2002;282:L468–L476. |
| 21. | Hind M, Corcoran J, Maden M. Alveolar proliferation, retinoid synthesizing enzymes, and endogenous retinoids in the postnatal mouse lung: different roles for Aldh-1 and Raldh-2. Am J Respir Cell Mol Biol 2002;26:67–73. |
| 22. | Massaro GD, Massaro D, Chambon P. Retinoic acid receptor-alpha regulates pulmonary alveolus formation in mice after, but not during, perinatal period. Am J Physiol Lung Cell Mol Physiol 2003;284:L431–L433. |
| 23. | Massaro GD, Massaro D, Chan WY, Clerch LB, Ghyselinck N, Chambon P, Chandraratna RA. Retinoic acid receptor-beta: an endogenous inhibitor of the perinatal formation of pulmonary alveoli. Physiol Genomics 2000;4:51–57. |
| 24. | McGowan S, Jackson SK, Jenkins-Moore M, Dai HH, Chambon P, Snyder JM. Mice bearing deletions of retinoic acid receptors demonstrate reduced lung elastin and alveolar numbers. Am J Respir Cell Mol Biol 2000;23:162–167. |
| 25. | Weibel ER. Morphometry of the human lung. Berlin: Springer-Verlag; 1963. |
| 26. | Massaro D, Teich N, Maxwell S, Massaro GD, Whitney P. Postnatal development of alveoli: regulation and evidence for a critical period in rats. J Clin Invest 1985;76:1297–1305. |
| 27. | Tschanz SA, Damke BM, Burri PH. Influence of postnatally administered glucocorticoids on rat lung growth. Biol Neonate 1995;68:229–245. |
| 28. | Massaro D, Massaro GD. Dexamethasone accelerates postnatal alveolar wall thinning and alters wall composition. Am J Physiol 1986;251:R218–R224. |
| 29. | Roth-Kleiner M, Berger TM, Tarek MR, Burri PH, Schittny JC. Neonatal dexamethasone induces premature microvascular maturation of the alveolar capillary network. Dev Dyn 2005;233:1261–1271. |
| 30. | Gudas L, Sporn MB, Roberts AB. Cellular biology and biochemistry of the retinoids. In: Sporn MB, Roberts AB, Goodman DWS, editors. The Retinoids: biology, chemistry and medicine. New York: Raven Press; 1994. pp. 443–520. |
| 31. | Rosa FW, Wilk AL, Kelsey FO. Teratogen update: vitamin A congeners. Teratology 1986;33:355–364. |
| 32. | Shenfelt RE. Morphogenesis of malformations in hamsters caused by retinoic acid: relation to dose and stage at treatment. Teratology 1972;5:103–118. |
| 33. | Beard RL, Chandraratna RAS. RAR-selective ligands: receptor subtype and function selectivity. In: Nau H, Blaner WS, editors. Retinoids: the biochemical and molecular basis of vitamin A and retinoid action. New York: Springer-Verlag; 1999. pp. 185–213. |
| 34. | Dawson MI, Zhang XK. Discovery and design of retinoic acid receptor and retinoid X receptor class- and subtype-selective synthetic analogs of all-trans-retinoic acid and 9-cis-retinoic acid. Curr Med Chem 2002;9:623–637. |
| 35. | Farmer LJ, Zhi L, Jeong S, Lamph WW, Osburn DL, Croston G, Flatten KS, Heyman RA, Nadzan AM. Retinoic acid receptor ligands based on the 6-cyclopropyl-2,4-hexadienoic acid. Bioorg Med Chem Lett 2003;13:261–264. |
| 36. | Martin B, Bernardon JM, Cavey MT, Bernard B, Carlavan I, Charpentier B, Pilgrim WR, Shroot B, Reichert U. Selective synthetic ligands for human nuclear retinoic acid receptors. Skin Pharmacol 1992;5:57–65. |
| 37. | Nagpal S, Chandraratna RA. Recent developments in receptor-selective retinoids. Curr Pharm Des 2000;6:919–931. |
| 38. | Piu F, Gauthier NK, Olsson R, Currier EA, Lund BW, Croston GE, Hacksell U, Brann MR. Identification of novel subtype selective RAR agonists. Biochem Pharmacol 2005;71:156–162. |
| 39. | Idres N, Marill J, Chabot GG. Regulation of CYP26A1 expression by selective RAR and RXR agonists in human NB4 promyelocytic leukemia cells. Biochem Pharmacol 2005;69:1595–1601. |
| 40. | Pendino F, Dudognon C, Delhommeau F, Sahraoui T, Flexor M, Bennaceur-Griscelli A, Lanotte M, Segal-Bendirdjian E. Retinoic acid receptor alpha and retinoid-X receptor-specific agonists synergistically target telomerase expression and induce tumor cell death. Oncogene 2003;22:9142–9150. |
| 41. | Sun SY, Yue P, Mao L, Dawson MI, Shroot B, Lamph WW, Heyman RA, Chandraratna RA, Shudo K, Hong WK, et al. Identification of receptor-selective retinoids that are potent inhibitors of the growth of human head and neck squamous cell carcinoma cells. Clin Cancer Res 2000;6:1563–1573. |
| 42. | Boisvieux-Ulrich E, Le Pechon-Vallee C, Million K, Baeza-Squiban A, Houcine O, Guennou C, Reichert U, Marano F. Differential effects of several retinoid receptor-selective ligands on squamous differentiation and apoptosis in airway epithelial cells. Cell Tissue Res 2000;300:67–81. |
| 43. | Corcoran J, Shroot B, Pizzey J, Maden M. The role of retinoic acid receptors in neurite outgrowth from different populations of embryonic mouse dorsal root ganglia. J Cell Sci 2000;113:2567–2574. |
| 44. | Elmazar MM, Ruhl R, Reichert U, Shroot B, Nau H. RARalpha-mediated teratogenicity in mice is potentiated by an RXR agonist and reduced by an RAR antagonist: dissection of retinoid receptor-induced pathways. Toxicol Appl Pharmacol 1997;146:21–28. |
| 45. | Elmazar MM, Ruhl R, Nau H. Synergistic teratogenic effects induced by retinoids in mice by coadministration of a RARalpha- or RARgamma-selective agonist with a RXR-selective agonist. Toxicol Appl Pharmacol 2001;170:2–9. |
| 46. | Kochhar DM, Jiang H, Penner JD, Beard RL, Chandraratna RAS. Differential teratogenic response of mouse embryos to receptor selective analogs of retinoic acid. Chem Biol Interact 1996;100:1–12. |
| 47. | Willhite CC, Dawson MI, Reichert U. Receptor-selective retinoid agonists and teratogenic activity. Drug Metab Rev 1996;28:105–119. |
| 48. | Elmazar MMA, Reichert U, Shroot B, Nau H. Pattern of retinoid-induced teratogenic effects: possible relationship with relative selectivity for nuclear retinoid receptors RARa, RARb, and RARg. Teratology 1996;53:158–167. |
| 49. | Maden M, Summerbell D, Maignan J, Darmon M, Shroot B. The respecification of limb pattern by new synthetic retinoids and their interaction with cellular retinoic acid-binding protein. Differentiation 1991;47:49–55. |
| 50. | Fujii H, Sato T, Kaneko S, Gotoh O, Fujii-Kuriyama Y, Osawa K, Kato S, Hamada H. Metabolic inactivation of retinoic acid by a novel P450 differentially expressed in developing mouse embryos. EMBO J 1997;16:4163–4173. |
| 51. | Tahayato A, Dolle P, Petkovich M. Cyp26C1 encodes a novel retinoic acid-metabolizing enzyme expressed in the hindbrain, inner earm first branchial arch and tooth buds during murine development. Gene Exp Patt 2003;3:449–454. |
| 52. | White JA, Guo Y-D, Baetz K, Beckett-Jones B, Bonasoro J, Hsu KE, Dilworth FJ, Jones G, Petkovich M. Identification of the retinoic acid-inducible all-trans-retinoic acid 4-hydroxylase. J Biol Chem 1996;271:29922–29927. |
| 53. | Nikawa T, Schulz WA, van den Brink CE, Hanusch M, van der Saag P, Stahl W, Sies H. Efficacy of all-trans-beta-carotene, canthaxanthin, and all-trans-, 9-cis-, and 4-oxoretinoic acids in inducing differentiation of an F9 embryonal carcinoma RAR beta-lacZ reporter cell line. Arch Biochem Biophys 1995;316:665–672. |
| 54. | Pijnappel WWM, Hendricks HFJ, Folkers GE, van den Brink CE, Dekker EJ, Edelenbosch C, van der Sagg PT, Durston AJ. The retinoid ligand 4-oxo-retinoic acid is a highly active modulator of positional specification. Nature 1993;366:340–344. |
| 55. | Kostetskii I, Yuan S-Y, Kostetskaia E, Linask KK, Blanchet S, Seleiro E, Michaille JJ, Brickell P, Zile M. Initial retinoid requirement for early avian development coincides with retinoid receptor coexpression in the precardiac fields and induction of normal cardiovascular development. Dev Dyn 1998;213:188–198. |
| 56. | Reijntjes S, Blentic A, Gale E, Maden M. The control of morphogen signalling: regulation of the synthesis and catabolism of retinoic acid in the developing embryo. Dev Biol 2005;285:224–237. |
| 57. | Reynolds NJ, Fisher GJ, Griffiths CE, Tavakkol A, Talwar HS, Rowse PE, Hamilton TA, Voorhees JJ. Retinoic acid metabolites exhibit biological activity in human keratinocytes, mouse melanoma cells and hairless mouse skin in vivo. J Pharmacol Exp Ther 1993;266:1636–1642. |
| 58. | Gaemers IC, van Pelt AMM, van der Saag PT, de Rooij DG. All-trans-4-oxo-retinoic acid: a potent inducer of in vivo proliferation of growth-arrested A spermatogonia in the vitamin A-deficient mouse testis. Endocrinology 1996;137:479–486. |
| 59. | Idres N, Benoit G, Flexor MA, Lanotte M, Chabot GG. Granulocytic differentiation of human NB4 promyelocytic leukemia cells induced by all-trans retinoic acid metabolites. Cancer Res 2001;61:700–705. |
