Clinical Research |
From the Division of Pulmonary Sciences and Critical Care Medicine (M.W.G., M.M., T.G., M.E.Y., L.A., H.G., B.G., R.M.T., N.F.V.), Pulmonary Hypertension Center, Department of Pathology, University of Colorado Health Sciences Center, Denver, Colo, and Division of Pulmonary and Critical Care Medicine (J.E.L.), Vanderbilt University Medical Center, Nashville, Tenn.
Correspondence to N.F. Voelkel, MD, Division of Pulmonary Sciences and Critical Care Medicine, 4200 E Ninth Ave, C272, Denver, CO 80262. E-mail norbert.voelkel{at}uchsc.edu
| Abstract |
|---|
|
|
|---|
Key Words: primary pulmonary hypertension familial primary pulmonary hypertension microarray gene profiling lung tissue
| Introduction |
|---|
|
|
|---|
A genetic predisposition is most certainly required
for the development of severe PH because (1) the background incidence
of PPH in the general population is very small; (2) PPH is familial
(FPPH) in
6% of the cases; and (3) the incidence in known at-risk
populationsfor example, in patients infected with the AIDS virus or
in women who have been treated with appetite-suppressant drugsis also
very
low.4 5 6 7 8
Progress in understanding the pathobiology of severe pulmonary hypertensive diseases has been hampered by 2 vexing major problems. First, there continues to be a lack of information regarding the natural history and the development of the critical vascular lesions. Second, there are no animal models that fully replicate the human pulmonary vascular pathology.
Although the pulmonary vascular lesions, in particular the so-called plexiform lesions, have in recent years been better characterized using immune histology and in situ hybridization technology,5 9 10 much more needs to be learned about the molecular pathology of these complex lesions. Although FPPH has been linked to germline mutations of the bone morphogenic protein (BMP) receptor II (BMPRII),11 12 and somatic mutations occur in plexiform lesion endothelial cells in sporadic PPH,13 it is not clear whether FPPH is molecularly distinct from sporadic, nonfamilial PPH.
The recent availability of gene microarray technology14 now permits the analysis of the gene expression profile of lung tissue obtained from patients with PPH and the comparison of the gene expression profile in the diseased lungs with that found in normal lung tissue. The output and density of data provided by this approach overcome the limitations of analyses based on the exploration of changes of a single gene at a time. We believe that lung tissue gene expression profiling is useful, because the diseased tissue contains information relevant to the disease process. Our present study was designed to analyze the gene expression pattern using high-quality RNA extracted from lungs of patients with PPH, including 2 patients with FPPH. We wished to address the main question, whether there is a molecular signature that distinguishes lung tissue of PH patients from structurally normal lung tissue. In addition, we questioned whether there is a gene expression pattern that distinguishes the lungs of patients with sporadic PPH from lungs of patients with FPPH. Here we report the significant differences in expressed genes between normal lungs and PPH lungs and the characteristic gene expression pattern that distinguishes PPH from FPPH.
We provide data that illustrate an expression pattern in PPH lungs indicative of an imbalance between genes regulating cell growth and genes regulating apoptosis. We also report that there are alterations in the lung tissues from patients with sporadic PPH in the expression of several genes coding for proteins of the transforming growth factor (TGF)-ß signaling superfamily.15 These data suggest that in addition to gene mutations of the BMPRII and TGF-ß receptor II (TGF-ßRII) genes,11 12 13 there are also changes in the level of expression of related TGF-ß family genes.15
| Materials and Methods |
|---|
|
|
|---|
|
Details for RNA preparation, reverse transcription, labeling, hybridization, and data analysis can be found in the online data supplement available at http://www.circresaha.org.
Data Analysis
Detailed protocols for data analysis of
Affymetrix microarrays and extensive documentation of the sensitivity
and quantitative aspects of the method have been
described.14 16 17
Concordant gene expression was established by comparing the expression level (for each gene) in a patient tissue sample with the expression level in 6 normal lung tissue samples. Because there were 6 PH patient samples, this allowed 36 comparisons.
To determine reproducibility of results, we obtained parallel duplicate lung RNA preparations and compared differences in Affymetrix arrays. We found false-positive changes in 1.7% to 3% of all genes analyzed. When all normal lung samples were compared with all diseased lung samples (6x6 or 36 independent comparisons), the average number of genes called changed per comparison was 1099 (18.1%). These data indicate that the changes between normal and diseased tissue gene expression are real differences.
Statistical Analysis
Raw data from array scans were averaged across all
gene probes on each array, and a scaling factor was applied to bring
the average intensity for all probes on the array to 2500. This allows
any sample to be normalized for comparison with any other comparable
sample, ie, lung versus lung. We used a paired, 2-tailed
t test to assess whether gene
expression differences were significant
(P<0.05).
| Results |
|---|
|
|
|---|
|
Difference in the Number of Genes Expressed in
PPH and Normal Lungs
From a total of
6800 genes assayed, 307 genes were
differently expressed (P<0.05)
when comparison was made between the normal lung tissue RNA (n=6) and
the PPH lung tissue RNA samples (n=6). Of these genes with changed
expression, 133 genes showed upregulation and 174 downregulation in
PPH.
Figure 2
shows a dendogram of the gene expression
patterns in the lungs from the 6 patients and the 6 normal lungs. The
expression pattern was similar for the 3 sporadic PPH lungs and
differed from that of the normal lungs, because the overall gene
expression patterns of the normal lungs resembled each other more
closely, as did those of the PPH lungs. This can be easily seen from
the clustering and abundance of the red and blue bars (higher and lower
degrees of gene expression, respectively), in
Figure 2
. The lung tissue gene expression pattern of the
female patient 6
(Table 1
) segregated with the 2 known cases of FPPH. This
patient, with extreme pulmonary artery pressure elevation, had
been orphaned, and her family history was unknown.
Table 2
provides a list of genes with a high degree
of concordance in their differential expression (PPH versus normal).
The list ranks those genes for which from 25 to 31 of 36 comparisons
were directionally concordant, ie, showed increased or decreased
expression. The highest concordance was found for the gene coding for
the necdin-related protein, which was decreased in PPH
(P<0.005) (31 of 36
comparisons). The necdin gene codes for a nuclear protein, which is
maternally imprinted and interacts with the transcription factor E2F1.
Disruption of the mouse necdin gene results in early postnatal
lethality.18
|
|
Analysis of Gene Expression by Cluster
Analysis
Comparison between normal lungs (n=6), the PPH lungs
(n=3), and the FPPH patients lungs (n=2) based on cluster
analysis showed that there were several clusters of reduced
gene expression that characterized PPH patient samples when compared
with FPPH patient samples and normal lung samples
(Figure 3
). Because patient 6 could not be separated from the
2 known cases of FPPH, also by cluster analysis, we tentatively
assigned this patient the symbol (F)PPH. Among the overexpressed genes
were genes coding for the following: notch homologue 3
(P<0.0001), cGMP-dependent
protein kinase (P<0.0002),
adenosine kinase
(P<0.002), a large number of
ribosomal proteins (altogether 26 of the 46 contained on the array),
the gene encoding thioredoxin
(P<0.006), cDNA topoisomerase
II
(P<0.02), and the
intracellular chloride channel 1
(P<0.02) (see also
Figure 3B
). There were overall more clustered and
nonclustered genes
(Tables 2
and 3
and online data supplement available
at http://www.circresaha.org), which showed more genes underexpressed
than overexpressed
(Figure 4
). These underexpressed genes can be broadly
categorized as genes coding for proteins involved in signal
transduction (mitogen-activated protein kinase K7
[P<0.01]), cell cycle
control (cyclin-dependent kinase 7), transcription and replication
factors, binding proteins (fatty acid binding protein 3 and actin
binding protein 278), enzymes (manosidase
type II and alcohol
dehydrogenase 5), and receptors (inositol 1,4,5-triphosphate receptor
types 1, 2, and 3)19 and
Tek, which is involved in angiogenesis and vessel
maintenance.20 21 22
The gene encoding gas 6 ("growth arrestspecific 6") was
underexpressed in all 6 patients with PH
(Figure 4
).
|
|
|
Surprisingly, cluster analysis demonstrated that the
lung tissue gene expression pattern of the FPPH lungs resembled the
pattern of the normal lungs more than it resembled the pattern
expressed in sporadic PPH lungs, and there clearly were genes that were
differently expressed in sporadic PPH when compared with FPPH lung
tissue
(Figures 4B
through F). Examples of these are the genes
encoding a voltage-gated, shaker-related potassium channel; caspase 9
eukaryotic translation elongation factor
1
23 ; the laminin receptor
124 ; the genes encoding an
inward rectifying K+ channel; the
endothelial PAS domain protein
125 ; the jun D
proto-oncogene; and the gene encoding
BMP4.26 We confirmed the
differences in gene expression between normal and PPH lung tissue for
the following genes using quantitative polymerase chain reaction:
ß-actin, laminin receptor 1, and several homeobox
genes27 (data not
shown).
| Discussion |
|---|
|
|
|---|
Altered Pattern of Expressed Genes
Related to Protein Synthesis and Degradation
The high concordance and clustering of
genes encoding multiple components of large supramolecular complexes,
which are likely organized by
synexpression28 such as the
proteasome and the ribosomal machinery, and further the clustered
differential expression of functionally related genes such as genes
coding for ion channels, ubiquitin-conjugating enzymes, or genes coding
for several mitochondrial proteins
(Table 3
), indicate that the identified differences between
normal and diseased lung tissue are real differences and do not
represent procedural artifacts (see also online data supplement
available at http://www.circresaha.org). We postulate that these
differences in the pattern of overexpressed, underexpressed, and absent
(nondetectable) genes between normal and diseased lung tissue can serve
as a platform to further explore the relevant elements of the
pulmonary hypertensive pathobiology. Abnormal cell growth and
phenotypic alterations of pulmonary arterial
endothelial and vascular smooth muscle cells are
without doubt elements of the pathobiology of
PPH.6 9 29
Overexpression of genes coding for ribosomal proteins
(Figure 3
) and underexpression of genes coding for protein
phosphatases (online data supplement available at
http://www.circresaha.org) and coding for a variety of ion channels
(Table 3
,
Figure 4
) point toward a broad disturbance in
protein synthesis, protein degradation, and altered vascular
reactivitycertainly in keeping with dysfunctional vascular
endothelial and smooth muscle cell phenotypes.
The proteasome is an essential component of the ATP-dependent
proteolytic pathway leading to the complete breakdown of proteins to
small peptides, elimination of abnormal proteins, and generation of
antigenic peptides that are being presented on major
histocompatability class I molecules to
lymphocytes.30 31
The ubiquitin-proteasomemediated proteolysis is also an important
mechanism that controls the destruction of cellular regulatory proteins
including the cyclins, p27, the transcription factor E2F, and the
receptors for epidermal growth factor and platelet-derived growth
factor.30
Altered Pattern of Genes Expressed in Muscle
and Endothelial Cells
The abnormal vascular cell phenotype may be related to
the decreased expression of several cytoskeletal and extracellular
matrix proteins. Genes involved in the control of endothelial cell
biology are differentially expressed between PPH and normal lung. For
additional details, see the online data supplement available at
http://www.circresaha.org.
Altered Pattern of Genes Involved in Cell
Growth and Apoptosis
Our decision to restrict this first gene microarray
study of PH to lung tissue from patients with PPH was guided by our
bias that patients with secondary PH constitute a less homogenous group
and by our knowledge that endothelial cell
proliferation in sporadic PPH is monoclonal, but polyclonal in
secondary PH,32 and that the
monoclonal endothelial cell growth may mechanistically
differ from the endothelial cell growth in secondary
PH.13 Because the
lumen-obliterating endothelial cell growth in sporadic
PPH shares features with a neoplastic process, ie, microsatellite
instability and mutations in the TGF-ßRII and Bax
genes,13 we wondered whether
the microarray analysis would shed further light on this
postulated mechanism. Although the lung tissue gene expression pattern
analysis cannot provide mechanistic evidence for such a
hypothesis, it perhaps provides additional, unexpected information in
support of a neoplasia-like cell growth program and molecular
information that transcends the description of altered
pulmonary vascular cell phenotypes. Cell growth in PPH
lungs may be facilitated by altered expression of cell cycle and
apoptosis-regulating genes. We recently found that there are
fewer terminal
deoxynucleotidyltransferasemediated
dUTP nick-end labeling (TUNEL)positive cells in lung tissue samples
from PPH patients when compared with normal adult lung tissue samples
and that there is an absence of TUNEL-positive cells in the plexiform
lesions.33 In this context,
it is of interest that the gene encoding the inositol
1,4,5-triphosphate receptor type III, which is involved in
apoptosis
regulation,19 34
was decreased in its expression in all of the PH lung tissues
(Table 3
) and that the gene encoding caspase 9 was at a very
low expression level in the 3 sporadic PPH patient lungs. On the other
hand, the gene encoding "defender of cell death" (DAD1), which when
lost triggers apoptosis, was decreased in PPH
(Table 3
).
The following findings could further support a
neoplasia-like paradigm of sporadic
PPH35 : the decreased
expression of the gene coding for gas 6, the thyroid receptor
interactor (P<0.008) (nuclear
receptor) that interacts with the cAMP response element binding
protein; a decrease in the expression of the gene encoding the DNA
repair gene (mapped to the human chromosome 2q25 locus)
KU8036 in all 6 PH tissue
samples together with an overexpression of the laminin receptor 1
gene37 ; and a decreased
expression of the gene coding for the eukaryotic
translation elongation factor 1
1 in sporadic
PPH lung tissue
(Figure 4
).
Altered Pattern of Expressed Genes in
FPPH
As stated above, we found that the gene
expression pattern of patient 6 more closely resembled the expression
pattern of the 2 patients with known FPPH and further that several
genes were differentially expressed in the tissues of the FPPH patients
and the (F)PPH patient when compared with PPH lung tissue
(Figures 3
and 4
). In fact, it is surprising how different the
lung tissue gene expression in sporadic PPH is when compared with FPPH.
When we explored whether expression differences would exist in 14 genes
localized to the FPPH chromosomal gene site (2q31 to
2q32)38 and compared PPH and
FPPH samples, no statistically significant differences were found.
However, there were a number of other genes that were found to be
highly significantly different in their expression when PPH and
FPPH/(F)PPH lungs were compared. Increased in sporadic PPH versus
FPPH were the genes encoding the following: apolipoprotein CIII
(P<0.0002),
5-hydroxytryptamine receptor 1B
(P<0.0002), BMP4
(P<0.005), cathepsin D
(P<0.001), serum
response
factor39 40 41
(P<0.01),
2-macroglobulin
(P<0.0008), jun D
proto-oncogene (P<0.0007),
laminin receptor 1 (P<0.004),
and caspase 9 (P<0.002).
Decreased in PPH versus FPPH were the following: interleukin enhancer
binding factor I (P<0.001);
the gene encoding a voltage-gated, shaker-related
K+ channel; and the genes coding for
TGF-ßRII,
Smad126 42 43 44
(Table 4
), and BMP-2 (see also online data supplement
available at
http://www.circresaha.org).
|
Conclusion
We began our study of differential gene expression in
PPH with the assumption that sporadic PPH is a disease with typical and
dramatic histological features, which are sufficiently
distinct from those of the structurally normal lung but essentially
indistinguishable from those found in FPPH lungs. We expected that,
when we compared normal and PH lung tissue, there would be a very large
number of differentially expressed genes, and we found this to be true.
But we were surprised to find that only 307 genes were significantly
different in their expression when PH tissues were compared with
structurally normal lung tissues. The microarray analysis of
whole lung tissue samples is useful, because it revealed a clear
signature pattern of expressed genes in patients with sporadic PPH and
FPPH. A large group of genes encoding ribosomal, mitochondrial, and
cytoskeletal proteins and genes encoding ion channels and enzymes were
differentially expressed between PH and normal lungs. Several
transcription factor genes and genes related to cyclin-dependent
kinases were different in their expression, indicating, in the
aggregate, that the PH gene signature reflects a profound imbalance in
the control of genes involved in cell proliferation and
apoptosis. Although the beneficial treatment and survival
effect that has been reported for PPH patients treated with
prostacyclin45 46
could well be associated with increased or repressed expression of one
or several genes, prostacyclin treatment was not a variable that
accounted for the difference in the gene expression profile between
sporadic PPH and FPPH.
Our study was conducted using total RNA extracted from random lung tissue samples; we did not include for analysis information that is based on the exploration of laser-microdissected plexiform lesions, but we did compare mRNA and total RNA from the same samples. We found that the analysis of lung tissue mRNA provides less information than the analysis of total RNA samples (see online data supplement available at http://www.circresaha.org). However, our work presented here and conducted with whole-tissue total RNA indicates that there are striking differences in the molecular expression profiles between sporadic and familial PPH. These differences in expression profiles are complemented by independent gene mutation analysis. Only the plexiform lesions in the lungs from patients with sporadic PPH13 but not from FPPH lungs display mutations of the Bax and TGF-ß RII genes (M. Yeager, unpublished data, 2001). Taken together, the RNA expression data and the DNA mutation data13 lead to the conclusion that sporadic and familial PPH are mechanistically distinct.
| Acknowledgments |
|---|
| Footnotes |
|---|
| References |
|---|
|
|
|---|
is
a novel substrate of rho-associated kinase.
Biochem Biophys Res Commun. 2000;278:7278.[Medline]
[Order article via Infotrieve]
-actin CArG elements
coordinate formation of a smooth muscle cell-selective, serum response
factor-containing activation complex. Circ
Res. 2000;86:221232.This article has been cited by other articles:
![]() |
E. D. Michelakis, M. R. Wilkins, and M. Rabinovitch Emerging Concepts and Translational Priorities in Pulmonary Arterial Hypertension Circulation, September 30, 2008; 118(14): 1486 - 1495. [Full Text] [PDF] |
||||
![]() |
L. Moreno-Vinasco, M. Gomberg-Maitland, M. L. Maitland, A. A. Desai, P. A. Singleton, S. Sammani, L. Sam, Y. Liu, A. N. Husain, R. M. Lang, et al. Genomic assessment of a multikinase inhibitor, sorafenib, in a rodent model of pulmonary hypertension Physiol Genomics, April 21, 2008; 33(2): 278 - 291. [Abstract] [Full Text] [PDF] |
||||
![]() |
C. C. Lewis, J. Y. H. Yang, X. Huang, S. K. Banerjee, M. R. Blackburn, P. Baluk, D. M. McDonald, T. S. Blackwell, V. Nagabhushanam, W. Peters, et al. Disease-Specific Gene Expression Profiling in Multiple Models of Lung Disease Am. J. Respir. Crit. Care Med., February 15, 2008; 177(4): 376 - 387. [Abstract] [Full Text] [PDF] |
||||
![]() |
M. F. Ramos, M. W. Lame, H. J. Segall, and D. W. Wilson Smad Signaling in the Rat Model of Monocrotaline Pulmonary Hypertension Toxicol Pathol, February 1, 2008; 36(2): 311 - 320. [Abstract] [Full Text] [PDF] |
||||
![]() |
B. O. Meyrick, D. B. Friedman, D. D. Billheimer, J. D. Cogan, M. A. Prince, J. A. Phillips III, and J. E. Loyd Proteomics of Transformed Lymphocytes from a Family with Familial Pulmonary Arterial Hypertension Am. J. Respir. Crit. Care Med., January 1, 2008; 177(1): 99 - 107. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Sakao, L. Taraseviciene-Stewart, C. D. Cool, Y. Tada, Y. Kasahara, K. Kurosu, N. Tanabe, Y. Takiguchi, K. Tatsumi, T. Kuriyama, et al. VEGF-R blockade causes endothelial cell apoptosis, expansion of surviving CD34+ precursor cells and transdifferentiation to smooth muscle-like and neuronal-like cells FASEB J, November 1, 2007; 21(13): 3640 - 3652. [Abstract] [Full Text] [PDF] |
||||
![]() |
H. H. Patel, S. Zhang, F. Murray, R. Y. S. Suda, B. P. Head, U. Yokoyama, J. S. Swaney, I. R. Niesman, R. T. Schermuly, S. S. Pullamsetti, et al. Increased smooth muscle cell expression of caveolin-1 and caveolae contribute to the pathophysiology of idiopathic pulmonary arterial hypertension FASEB J, September 1, 2007; 21(11): 2970 - 2979. [Abstract] [Full Text] [PDF] |
||||
![]() |
F. A. Masri, W. Xu, S. A. A. Comhair, K. Asosingh, M. Koo, A. Vasanji, J. Drazba, B. Anand-Apte, and S. C. Erzurum Hyperproliferative apoptosis-resistant endothelial cells in idiopathic pulmonary arterial hypertension Am J Physiol Lung Cell Mol Physiol, September 1, 2007; 293(3): L548 - L554. [Abstract] [Full Text] [PDF] |
||||
![]() |
S. Bonnet, G. Rochefort, G. Sutendra, S. L. Archer, A. Haromy, L. Webster, K. Hashimoto, S. N. Bonnet, and E. D. Michelakis The nuclear factor of activated T cells in pulmonary arterial hypertension can be therapeutically targeted PNAS, July 3, 2007; 104(27): 11418 - 11423. [Abstract] [Full Text] [PDF] |
||||
![]() |
J. U. Raj, C. Aliferis, R. M. Caprioli, A. W. Cowley Jr., P. F. Davies, M. W. Duncan, D. J. Erle, S. C. Erzurum, P. W. Finn, H. Ischiropoulos, et al. Genomics and proteomics of lung disease: conference summary Am J Physiol Lung Cell Mol Physiol, July 1, 2007; 293(1): L45 - L51. [Full Text] [PDF] |
||||
![]() |
A. Zakrzewicz, F. M. Kouri, B. Nejman, G. Kwapiszewska, M. Hecker, R. Sandu, E. Dony, W. Seeger, R. T. Schermuly, O. Eickelberg, et al. The transforming growth factor-{beta}/Smad2,3 signalling axis is impaired in experimental pulmonary hypertension Eur. Respir. J., June 1, 2007; 29(6): 1094 - 1104. [Abstract] [Full Text] [PDF] |
||||
![]() |
Y. J. Suzuki, H. Nagase, C. M. Wong, S. V. Kumar, V. Jain, A.-M. Park, and R. M. Day Regulation of Bcl-xL Expression in Lung Vascular Smooth Muscle Am. J. Respir. Cell Mol. Biol., June 1, 2007; 36(6): 678 - 687. [Abstract] [Full Text] [PDF] |
||||
![]() |
G. Hansmann, R. A. Wagner, S. Schellong, V. A. de Jesus Perez, T. Urashima, L. Wang, A. Y. Sheikh, R. S. Suen, D. J. Stewart, and M. Rabinovitch Pulmonary Arterial Hypertension Is Linked to Insulin Resistance and Reversed by Peroxisome Proliferator-Activated Receptor-{gamma} Activation Circulation, March 13, 2007; 115(10): 1275 - 1284. [Abstract] [Full Text] [PDF] |
||||
![]() |
A. M. Assaad, S. M. Kawut, S. M. Arcasoy, E. B. Rosenzweig, J. S. Wilt, J. R. Sonett, and A. C. Borczuk Platelet-Derived Growth Factor Is Increased in Pulmonary Capillary Hemangiomatosis Chest, March 1, 2007; 131(3): 850 - 855. [Abstract] [Full Text] [PDF] |
||||
![]() |
T. M. Bull, C. D. Coldren, M. W. Geraci, and N. F. Voelkel Gene Expression Profiling in Pulmonary Hypertension Proceedings of the ATS, January 1, 2007; 4(1): 117 - 120. [Abstract] [Full Text] [PDF] |
||||
![]() |
J.-F. Jasmin, I. Mercier, J. Dupuis, H. B. Tanowitz, and M. P. Lisanti Short-Term Administration of a Cell-Permeable Caveolin-1 Peptide Prevents the Development of Monocrotaline-Induced Pulmonary Hypertension and Right Ventricular Hypertrophy Circulation, August 29, 2006; 114(9): 912 - 920. [Abstract] [Full Text] [PDF] |