Transcript signatures of lymphocytic bronchitis in lung allograft biopsy specimens

Abstract

Background: Rejection and obliterative bronchiolitis are barriers to sustained graft function in recipients of transplanted lungs. Early detection is hindered by inadequate tests and an incomplete understanding of the molecular events preceding or accompanying graft deterioration.

Methods: Hypothesizing that genes involved in immune responses and tissue remodeling produce biomarkers of rejection, we measured the expression of 192 selected genes in 72 sets of biopsy specimens from human lung allografts. Gene transcripts were quantified using a 2-step, multiplex, real-time polymerase chain reaction approach in endobronchial and transbronchial biopsy specimens from transplant recipients without acute infections undergoing routine surveillance bronchoscopy.

Results: Comparisons of histopathology in parallel biopsy specimens identified 6 genes correlating with rejection as manifested by lymphocytic bronchitis, a suspected harbinger of obliterative bronchiolitis. For example, beta2-defensin and collagenase transcripts in inflamed bronchi increased 37-fold and 163-fold, respectively. By contrast, these transcripts did not correlate with acute rejection in transbronchial specimens. Further, no correspondence was noted between histopathologic bronchitis and parenchymal rejection when endobronchial and transbronchial samples were obtained from the same patient.

Conclusions: Our highly sensitive method permits quantitation of many gene transcripts simultaneously in small, bronchoscopically acquired biopsy specimens of allografts. Transcript signatures obtained by this approach suggest that airway and alveolar responses to rejection differ and that endobronchial biopsy specimens assess lymphocytic bronchitis and chronic rejection but are not proxies for transbronchial biopsy specimens. Further, they reveal changes in airway expression of the specific genes involved in host defense and remodeling and suggest that the measurement of transcripts correlating with lymphocytic bronchitis may be diagnostic adjuncts to histopathology.

Figure 1

Correlation of gene transcript levels with histopathology in endobronchial biopsy (EBB) and transbronchial biopsy (TBB) specimens. Results for selected pairs of genes are shown in each panel. Transcript levels were determined by multiplex, real-time polymerase chain reaction (PCR). For all panels the y-axis unit is the log₁₀ value of the relative copy number, which is the transcript level of the gene of interest normalized to that of the glyceraldehyde 3-phosphate dehydrogenase (GAPDH) gene in the sample. All data are plotted on the same scale to facilitate comparison. Each data point is the mean ± standard error of the relative copy number of the transcript in patients with a given histopathologic grade as determined in parallel biopsy samples obtained from the same patient at the same sitting. For EBB specimens, the scale ranges from B0 (no lymphocytic airway inflammation) to B4 (severe inflammation with epithelial dissociation, ulceration and necrosis). B3 is the most severe grade in these subjects. For TBB, the scale ranges from A0 (normal parenchyma) to A4 (severe acute rejection with diffuse perivascular, interstitial and air space infiltrates of mononuclear cells). The highest grade in these biopsies is A3 (1 patient), which is grouped with A2 (2 patients). The genes profiled here are collagenase (matrix metalloproteinase-1, MMP1), MMP2, tissue inhibitor of metalloproteinases (TIMP) 1 and 2, β2-defensin (DEFB2), lung chemokine SCYA18, fibronectin (FN1) and collagen α chain (COL1A1). Results of statistical tests applied between a given data point and the value at B0 or A0 are indicated (*p < 0.05; **p < 0.01; ****p < 0.001).

Figure 2

Transcript profiles in paired endobronchial biopsy (EBB) and transbronchial biopsy (TBB) specimens. The position of each data point is determined by transcript level (copy number relative to level of glyceraldehyde 3-phosphate dehydrogenase [GAPDH] transcript) of a given gene product in EBB and TBB specimens obtained from the same patient at the same sitting. In each graph, the diagonal line marks where data points will fall if expressed at identical levels in both specimens of a given EBB/TBB specimen pair. Data points representing genes expressed preferentially in EBB or TBB specimens fall above and below the line, respectively. Examples are grouped according to relationships between gene products. The upper 6 panels display results involving selected matrix metalloproteinases (MMPs, MMP1, MMP2, MMP9, MMP11) and tissue inhibitor of metalloproteinases (TIMPs). Note the large differences in expression patterns between products, with MMP1 exhibiting a wide range of expression compared with TIMPs, which are near-constantly expressed at higher levels than any MMP. MMP9 is an example of a gene that is strongly overexpressed in TBB relative to EBB specimens, whereas the opposite is the case for MMP11. Among mucin genes depicted in the third row, MUC4 and MUC5AC are strongly overexpressed in EBB specimens. However, this is not true of all mucin genes, for MUC1 transcripts, while varying about 100-fold, are expressed at near-equivalent levels in any given EBB/TBB pair and a strongly correlated (r = 0.816, p < 0.001). Other strongly correlated genes are shown in the bottom row, including PRSS22 (ε-tryptase), MIF (macrophage migration inhibitory factor) and CTGF (connective tissue growth factor), with correlation coefficients of 0.862, 0.800, and 0.829, respectively (each with p < 0.001). These examples underscore the diversity of the expression of transcripts in terms of magnitude, range, and biopsy site.