The role of Smad3 -signalling in the regulation of lung homeostasis and in the pathophysiology of respiratory diseases such as allergic asthma is poorly understood. It has been shown, however, that TGF-β is present in the airways and that the levels of certain TGF-β isoforms have been shown to be elevated in asthmatic airways. Another member of the TGF superfamily, activin A, has also been suggested to be involved in the pathogenesis of asthma 12, 13
. Since Smad3 is a critical mediator of TGF-β -signalling and activin A signalling, we utilized Smad3 deficient mice to explore the role of Smad3 -signalling in the regulation of lung functions both in the healthy situation as well as in the diseased state.
Lung histology from naive Smad3-/- mice revealed that the lack of TGF-β -signalling caused a marginal infiltrate of lymphocytes and occasionally modest increases in mucin production in the bronchial epithelium. These changes were accompanied by significant increases in the expression of Th2 cytokine IL-4 and GATA-3, an important transcription factor regulating IL-4 expression. Similarly IL-5 was also significantly increased although the absolute level remained low. The levels of IL-13 were almost nonexistent. It has been shown that GATA-3 expression is suppressed by TGF-β 23
. Thus, the increase of GATA-3 and IL-4 could be viewed to result from the release from TGF-β suppression of GATA-3. In addition to the increase in Th2 cytokines, we found a significant elevation in the expression of proinflammatory cytokines and chemokines (IL-1β and MIP-1α, respectively), the latter being indicative of macrophage activation. In line with this finding, a tendency towards activation of macrophages with regard to matrix metalloproteinases has been observed by Bonniaud et al.
in Smad3 knockout mice 24
. The expression levels of a regulatory cytokine, IL-10, were completely abolished in naive Smad3 deficient mice compared to WT mice. The baseline levels of IL-10 in the lungs of WT mice are, however, quite low and it is therefore unclear whether this additional reduction could lead to the increased expression of proinflammatory cytokines and IL-4 seen. The reason for the decrease of IL-10 as well as the reasons for the low levels of IL-5 and IL-13 could be that the presence of Smad3 is required for GATA3 regulation 25
, this is discussed in association with the asthma model.
Our results suggest that Smad3 signalling is important in the suppression of the proinflammatory as well as the IL-4 response in the airways in the healthy condition, it should be noted that the suppressed features mentioned are also associated with asthma.
When studying the lungs of mice exposed to allergen in the asthma model we found that there was a pronounced increase in the amount of mucin in both PBS and OVA treated Smad3-/- mice compared to WT mice. Our results also revealed that airway hyperreactivity to inhaled metacholine, a classical hallmark of allergic asthma, was significantly increased in OVA sensitized Smad3-/- mice. In contrast, no significant differences between OVA and PBS sensitized mice were found in WT mice. These results probably underestimate the actual degree of AHR since C57BL/6 mice (strain used in the present study) are more resistant to bronchoconstrictors than other mouse strains, and differences between mice might be difficult to observe using whole body plethysmography 24
. Indeed, we were able to observe signs of respiratory distress in Smad3-/- mice after allergen exposure but no symptoms were visible in WT mice. Taken together, we found that the lack of Smad3-/- signalling led to an exaggerated asthmatic phenotype in allergic asthma model. The exaggerated asthmatic phenotype seen in this work resembles that seen in the work of Nakao et al 18
where expression of the inhibitory Smad7 was used to disturb signalling.
Naive Smad3 deficient mice demonstrated elevated levels of IL-4 protein in their lungs. In line with this, the levels of IL-4 mRNA and protein in the airways of PBS treated Smad3-/- mice were significantly increased compared to PBS treated WT mice. However, the IL-4 levels were equivalently increased in OVA sensitized Smad3-/- and WT mice. These results suggest that cytokines belonging to the TGF superfamily play an important role in the regulation of IL-4 production in the normal state but their significance is less critical during inflammation induced in this case by immunization with OVA allergen and/or adjuvant. In contrast to IL-4 expression, the levels of IL-13 (another important Th2 type cytokine) were significantly decreased in OVA sensitized lungs of Smad3-/- mice compared to WT mice. It has been previously reported that GATA-3 interacts extensively and apparently directly with Smad3 to regulate GATA target genes, including IL-10 and IL-5, through GATA3 binding sites in the promoter regions 25
. Thus the low level of increase of IL-5 and the decrease of IL-10 in naive Smad3-/- mice could be caused by an inhibition of GATA-3 function by the lack of Smad3. IL-13 is regulated through a GATA-3 binding promoter region 26
, but there are no earlier reports that its expression would be dependent on Smad3 25
. Our findings, interestingly, suggest a critical dependence of GATA-3 for Smad3 in regulating IL-13 expression. In contrast, no GATA binding exists in the IL-4 promoter and GATA-3 regulates IL-4 production through distal regulatory elements 26-28
. In the asthma model, in contrast to naive mice, there were no significant differences in the amounts of GATA-3 mRNA levels between Smad3-/- and WT. Furthermore, IL-10 levels in the asthma model were much higher than in the naive state and they did not reveal any significant differences between Smad3-/- and WT mice mice, suggesting that there is a difference in the regulation of GATA3 and IL-10 cytokines in the naive state and the inflamed state. In the case of IL-10 it appears that in PBS exposed lungs other regulatory mechanisms override the lack of GATA3 and increase IL-10 mRNA levels substantially as compared to naive mice. Taken together, we think that the unexpectedly low levels or decreases of IL-5, IL-10, and IL-13 in various contexts can best be explained by a need for interactions between GATA3 and Smad3. However, depending on the situation, there are obviously other regulatory mechanisms that strongly affect the expression of these cytokines. Further studies are needed to clarify the molecular mechanisms for GATA3 and Smad3 effects on cytokine expression.
Increased amounts of IL-13 have been previously associated with increased airway hyperreactivity and mucus overproduction. It is interesting that in the present study Smad3 deficient mice with diminished IL-13 production clearly exhibited increased airway hyperreactivity and substantially increased mucus production. This result is in agreement with the finding of Perkins et al 29
demonstrating a role for IL-4 in promoting all the Th2 tissue responses independent of IL-13.
Our results are in many respects contradictory to those described by Le et al 30
.They showed that a Smad3 deletion led to a decrease in mucin production and eosinophilia whereas no changes was seen with the AHR. The probable reason for this discrepancy is that in the work of Le et al a different Smad3 deletion was used, i.e that described by Zhu et al in 1998 31
which has exons 2 deleted on both alleles. In contrast we have used a Smad3 deletion where exons 8 on both alleles were deleted 19
. It is well known that these different Smad3 deletions often give contradictory results. Perhaps the most interesting feature of the Smad3 (exon 2) deletion is a major increase in the incidence of colonic cancer. No increases in colonic cancer has been reported from the Smad3 (exon 8) deletions. As far as we know the reason for these discrepancies have not been fully determined, Yang et al 1999 19
has discussed this topic to some extent in his article and propose a possible explanation i.e. that some part of the mutant transcript (exons 4-9) could actually be translated in the Smad3 (exon 2) deletion. In contrast there is evidence showing that no part of the mutant transcript in the Smad3 (exon 8) deletion is translated. In addition there are differences in the genetic background of the mice but according to Yang et al this is not likely to explain the discrepancies between the different Smad3 deletions. Thus, it is possible that the explanation for the differences between our results and those of Le et al is that in the latter case a part of Smad3 is translated. If we now consider other models where TGF-β or its receptor has been manipulated the results are in many ways similar to the data concerning the inflammatory state presented here i.e. eosinophilia and AHR is increased and mucin production is increased when TGF β1 levels or effect are decreased 11, 18
. Thus we think that our results are much more likely to reflect a generalizable role for Smad3 signalling in asthma.
TGF-β is not entirely dependent on Smad dependent pathways for its signalling 32, 33
and a total lack of TGF-β will cause a much more severe phenotype than that seen in the Smad3-/- mice 2
. This study showed that a lack of Smad3 in naive mice caused an increase in the central Th2 transcription factor GATA-3 and changes in IL-4 and proinflammatory cytokine levels. Interestingly the lack of Smad3 and presumably TGF-β signalling led to an exaggerated asthmatic phenotype in an asthma model, in line with earlier findings. However, this exaggerated asthmatic phenotype was not accompanied by any increase in IL-4 or other Th2 cytokines, in contrast to earlier findings with other forms of inhibition of TGF-β 11, 18
in asthma models. This indicates that in the inflammatory state, TGF-β or related cytokines function as a counterbalance to IL-4 effects rather than as a regulator of its expression.