Differential epithelial expression of SHH and FOXF1 in usual and nonspecific interstitial pneumonia

https://doi.org/10.1016/j.yexmp.2005.12.003Get rights and content

Abstract

Morphogenetic factors have been shown to play a role in embryogenesis and post-embryonic disease. Interstitial pulmonary fibrosis is a chronic and often progressive disorder that can lead to end-stage cystic lung. Its two major subtypes, usual interstitial pneumonitis (UIP) and nonspecific interstitial pneumonitis (NSIP) differ in their response to immunosuppressive regimens, with UIP having a significantly worse prognosis. The clinical and histologic overlap between these disorders is substantial, and there are no ancillary findings that can accurately distinguish them. We examined surgical and autopsy specimens of lung in 13 cases of patients with either UIP or the fibrotic variant of NSIP (NSIP-F) for their expression of Sonic hedgehog (Shh) and Foxf1 in situ. We identified a pattern of strong Shh expression with weak expression of Foxf1 in all cases of UIP and a complementary expression of Shh and Foxf1 in cases of NSIP-F. We conclude that morphogenetic genes may participate differentially in the pathogenesis of UIP and NSIP-F.

Introduction

The idiopathic interstitial pneumonias (IIP) represent a set of diffuse parenchymal lung disorders with diverse histologies (Consensus, 2002). Idiopathic pulmonary fibrosis (IPF) constitutes a subset of IIP that generally affects middle-aged patients, yielding respiratory compromise. The most common variant, usual interstitial pneumonitis (UIP) is the most common or “usual” form of IPF. Whereas most cases of UIP are idiopathic, comparable histology can be seen in pulmonary disorders due to collagen-vascular disease, drug toxicities, or genetic abnormalities.

The lung in UIP shows variable degrees of interstitial lymphohistiocytic inflammation and diffuse, patchy, subpleural fibrosis. The scarring in UIP shows mature collagen with foci of active fibroplasia. Progressive scarring leads to the development of cystic airspaces lined by respiratory epithelium, i.e., end-stage “honeycomb” lung.

In 1994, Katzenstein and Fiorelli (1994) proposed the term nonspecific interstitial pneumonia (NSIP) to include the cases of IPF that do not exhibit the diagnostic histological features of UIP. NSIP accounts for 14–35% of biopsies in patients with IPF and can be subclassified into cellular and fibrotic types. Unlike UIP, the lungs in NSIP show diffuse temporally homogeneous cellular infiltration and scarring. Like UIP, NSIP can also lead to subpleural cyst formation.

Whereas the cellular variant of NSIP has a good prognosis, the prognosis of the fibrotic variant (NSIP-F) is on average better than for UIP (Monaghan et al., 2004). At times, it is can be impossible to distinguish NSIP-F from UIP based on clinical and radiographic features, and lung biopsy is required. But the pathologic diagnosis can be challenging, when biopsy material is limited and fibroblastic foci and cysts are present in NSIP-F. Flaherty and colleagues have demonstrated that one in four biopsy sites of IPF shows discordant results, i.e., with one site showing changes of UIP, while another shows typical NSIP-F (Flaherty et al., 2001). Katzenstein and colleagues demonstrated areas indistinguishable from NSIP in 19/20 lungs that otherwise showed classic UIP (Katzenstein et al., 2002).

Accurately distinguishing NSIP from UIP is important because the prognosis of NSIP and its response to immunosuppressive therapies are better than for UIP. Whereas rigorously defined UIP is virtually always progressive and often lethal, the disease-specific mortality rates for NSIP-F range from 11 to 68% with a mean of 33% (Riha et al., 2002). Whether part of the mortality statistics attributed to NSIP-F might represent diagnostic confusion with UIP is uncertain, as there are presently no reliable immunohistochemical or molecular markers that can accurately distinguish NSIP-F from UIP.

Factors important in embryological development participate in adult tissue homeostasis and disease. Genes recognized to play a critical role in embryogenesis, including TGF-β, integrins, and tenascin, have been found to be upregulated in chronic pulmonary fibrosis (Stewart et al., 2003). The hedgehog (hh) family of secreted proteins are morphogens critical in vertebrate development (Bitgood and McMahon, 1995, Chuang and McMahon, 1999). Sonic hedgehog (Shh) paracrine signaling pathways contribute specifically to branching morphogenesis in the embryonic lung (Bellucci et al., 1997, Hogan et al., 1997). Downstream factors in the pathway include the Shh receptors, Ptc and Smo, which activate downstream expression of other factors including transcription factors (Ingham, 2001, Cohen, 2003). The Fox family of forked winged helix transcription factors participate in embryogenesis, aging, and neoplasia (Katoh and Katoh, 2004) and is also a target of hh signaling. (Mahlapuu et al., 1998, Mahlapuu et al., 2001a, Mahlapuu et al., 2001b, Jeong et al., 2004, Katoh and Katoh, 2004). Of the Fox genes, Foxf1 is the most strongly expressed in the developing murine lung and has a role in its development (Mahlapuu et al., 2001b).

In human fetal lungs, FOXF1 is detected primarily in the mesenchyme surrounding the tips of the distal pulmonary epithelium, with modest expression by the proximal bronchiolar epithelium at 8 weeks of gestation. FOXF1 epithelial expression peaks at 16 weeks and persists until term, at which time FOXF1 is no longer detected in the lung. Foxf1 has not been reported in murine pulmonary epithelium, and this may reflect either a species difference or the failure to detect epithelial expression within a limited temporal window. Human pulmonary mesenchymal and epithelial expression is consistent with the hypothesis that Foxf1 is a downstream target for Shh signaling, as the human fetal expression of SHH and its receptor PTCH1 mirrors that of FOXF1 (manuscript in preparation).

As studies have found a role for the hh pathway in inflammatory and fibrotic diseases of the human gut (Nielsen et al., 2004) and lung (Stewart et al., 2003), we hypothesized that the hh pathway might be activated in IPF. We analyzed the mRNA expression patterns of SHH and FOXF1 in normal lung tissues and those diagnosed with IPF. We detected expression in a pattern that may help differentiate UIP from NSIP-F.

Section snippets

Materials and methods

Pulmonary tissues of 13 patients were examined with institutional IRB approval. Patients ranged from 35 to 81 years of age and included 7 men and 6 women. Sections of normal lung from two pulmonary resections for benign neoplasia served as controls. In patients with UIP or NSIP-F, concurrent diagnoses included pulmonary thromboembolic disease, cavitary tuberculosis, acquired immunodeficiency syndrome (AIDS), and pulmonary squamous cell carcinoma (Table 1). Specimens were obtained either by

Results

End-stage fibrosis with honeycomb cystic changes was identified in all cases of UIP and NSIP-F. The cysts were predominantly subpleural in both disorders. The cysts of UIP were lined predominantly by columnar or high cuboidal respiratory epithelium and often contained inspissated mucin (Fig. 1A). The cystic spaces of NSIP were lined predominantly by either cuboidal or low columnar epithelium and rarely showed substantial mucin (Fig. 1B). In addition, the honeycomb cysts of UIP were larger than

Discussion

In this study, we report the differential pulmonary expression of SHH and FOXF1 in lungs with UIP and NSIP-F. Specifically, SHH was expressed strongly in the epithelial lining of honeycomb cysts of UIP, whereas FOXF1 was preferentially expressed by the lining epithelium of NSIP-F, although the latter change did not achieve statistical significance, due to the small number of cases and to a single case that showed the same SHH/FOXF1 phenotype of UIP.

The radiographic identification of “honeycomb”

Acknowledgments

The authors wish to thank P. Carlsson for the reagents and insightful conversations and M. Lozano for the technical assistance. This research was funded in part by an NIH grant HD99-008 to DJR.

References (35)

  • P. Chuang et al.

    Vertebrate Hedgehog signaling modulated by induction of a hedgehog-binding protein

    Nature

    (1999)
  • M.M. Cohen

    Hedgehog signaling network

    Am. J. Med. Genet.

    (2003)
  • A.T.S.E.R.S.I.M. Consensus

    Classification of the idiopathic interstitial pneumonias:

    Am. J. Respir. Crit. Care Med.

    (2002)
  • C. Descipio et al.

    Subtelomeric deletions of chromosome 6p: molecular and cytogenetic characterization of three new cases with phenotypic overlap with Ritscher-Schinzel (3C) syndrome

    Am. J. Med. Genet., A

    (2005)
  • R.P. Erickson

    Forkhead genes and human disease

    J. Appl. Genet.

    (2001)
  • K. Flaherty et al.

    Histopathologic variability in usual and nonspecific interstitial pneumonia

    Am. J. Respir. Crit. Care Med.

    (2001)
  • B.L. Hogan et al.

    Branching morphogenesis of the lung: new models for a classical problem

    Cold Spring Harbor Symp. Quant. Biol.

    (1997)
  • Cited by (38)

    View all citing articles on Scopus
    View full text