Download Adrenal Medullary Tumors, Molecular and Cell Biology of

Molecular and Cellular Biology of Pheochromocytomas and Extra-adrenal
Paragangliomas
Arthur S. Tischler
Dept. of Pathology, Tufts New England Medical Center, Boston, MA
[email protected]
The 2004 WHO classification of endocrine tumors [1] defines
pheochromocytoma as a tumor arising from chromaffin cells in the adrenal medulla.
Closely related tumors in extra-adrenal sympathetic and parasympathetic paraganglia are
classified as extra-adrenal paragangliomas. A pheochromocytoma is an intra-adrenal
sympathetic paraganglioma. Although arbitrary, this nomenclature serves to emphasize
important distinctive properties of intra-adrenal tumors that must be taken into account in
clinical practice and research. Those include a lower rate of malignancy (~5% overall, vs
~20% for extra-adrenal paragangliomas associated with the sympathetic nervous system
[2]), an often adrenergic phenotype (extra-adrenal paragangliomas are almost always
noradrenergic [3]), and a proclivity to occur in association with particular genetic
disorders (particularly MEN2) [4]. It can be reasonably argued that this classification
may detract from efforts to understand the pathobiology of tumors that, overall, are more
similar than different. Pheochromocytomas and extra-adrenal paragangliomas often have
similar or identical morphology, share a mostly identical neuroendocrine phenotype, and
sometimes have the same genetic predisposition. Nonetheless, the need for consistency
dictates that the WHO nomenclature, while imperfect, should be adhered to. At the same
time, the biological basis for differences in genotype, phenotype and malignancy based
on anatomic site must be resolved.
Efforts to understand the pathobiology of pheochromocytoma/paraganglioma
have been spurred by recent advances in genetics and gene expression profiling. New
understanding of familial syndromes, especially von Hippel-Lindau disease (VHL) and
recently defined familial paraganglioma (PGL) syndromes, has provided a platform from
which to explore both familial and sporadic tumors. The current challenge is to relate
catalogs of genetic and phenotypic markers to tumor cell biology.
Genetics
Physicians have traditionally been taught to remember the clinical properties of
sympathetic paragangliomas according to the "10 per cent rule"- 10% familial, 10%
malignant, 10% extra-adrenal. That rule is no longer tenable. Both hospital- and
population-based studies have demonstrated occult germline mutations characteristic of
familial pheochromocytoma/paraganglioma syndromes in more than 20% of patients
presenting with apparently sporadic tumors [5,6], bringing the percentage of tumors with
a known genetic basis close to 30%. In addition, the anatomic locations of tumors and
their risk of malignancy vary according to the underlying genetic defect.
Hereditary syndromes long known to be associated with development of
pheochromocytomas/paragangliomas are Multiple Endocrine Neoplasia (MEN) 2A and
2B, von Hippel-Lindau disease (VHL) and neurofibromatosis type 1 (NF1) due
1
respectively to mutations of the RET proto-oncogene and the VHL and NF1 tumor
suppressor genes. The list is now expanded to include familial paraganglioma (PGL)
syndromes caused by mutations of succinate dehydrogenase genes SDHD (PGL1), SDHC
(PGL3), and SDHB (PGL4), which also appear to function as tumor suppressor genes
[7](syndrome numbers were assigned in order of discovery) (Table 1). In addition, VHL
is divided into types 1 and 2, defined by the absence or presence of susceptibility to
pheochromocytomas/paragangliomas (TABLE 2). Rates of malignacy range from <3%
for tumors with RET mutations to >50% for those with mutated SDHB [6].
TABLE 1. FAMILIAL PARAGANGLIOMA SYNDROMES
SYNDROME
PGL1
PGL3
PGL4
GENE
(chromosome)
TUMOR DISTRIBUTION *
SDHD (11q23)
SDHC (1q21-23)
SDHB (1p36)
ADRENAL
OTHER
SYMPATHETIC
++
++
++
++
PARASYMPATHETIC
++
++
++
*Adapted from WHO bluebook 2004 [1]. In contrast to PGL1 or 4, PGL3 tumors are relatively rare, seldom malignant and
seldom multifocal [8].
TABLE 2. SUBTYPES OF VHL DISEASE
MAJOR TUMOR DISTRIBUTION *
SUBTYPE
PHEOCHROMOCYTOMA
OR OTHER SYMPATHETIC
PARAGANGLIOMA
1
2A
2B
2C
√
√
√
RENAL CELL
CARCINOMA
HEMANGIOBLASTOMA
OR RETINAL ANGIOMA
√
√
√
√
√
Adapted from WHO bluebook 2004 [1]. Parasympathetic paragangliomas occur rarely in VHL disease. Note that
pheochromocytomas in VHL 2C occur without any other VHL stigmata
Somatic mutations of the genes responsible for hereditary
pheochromocytomas/paragangliomas are uncommon in tumors that are truly sporadic.
The latter are reported to harbor somatic mutations of RET in up to~ 10% of cases and
VHL mutations in ~ 4%. Somatic mutations of SDHB or SDHD have been reported
occasionally in sporadic tumors [9,10] that may be solitary or multiple [9]. A novel
2
aspect of PGL1 is a mode of transmission that appears to involve genomic imprinting,
i.e., tumors occur only after paternal transmission of the mutated gene [7].
Several additional additional unknown genes apparently confer susceptibility to
hereditary pheochromocytomas/paragangliomas in some kindreds. One of two recently
identified novel gene loci is on chromosome 2q and appears to be the site of a tumor
suppressor gene. Another, on chromosome 16p1, is at or near the locus for hereditary
neuroblastoma [11]. Synergy between 2q and 16p is a model suggested to explain
inherited tumor susceptibility in some patients. In contrast to other
pheochromocytoma/paraganglioma syndromes, tumor susceptibility involving these loci
is apparently transmitted as a recessive trait [11].
Sympathetic paragangliomas occur anywhere in the distribution of the
sympathetic nervous system from the superior cervical ganglion to the walls of pelvic
organs. The distribution of parasympathetic paragangliomas is limited to cranial, cervical
and thoracic branches of the vagus and glossopharyngeal nerves. Tumors in patients with
MEN2 or neurofibromatosis often arise in or near the adrenal. In contrast, VHL tumors,
while predominantly intra-adrenal, are not uncommon in extra-adrenal locations and
tumors associated with SDH mutations are predominantly extra-adrenal [6]. SDHD or
SDHC mutations are usually confined to parasympathetic paragangliomas in the head and
neck, while the combination of sympathetic and parasympathetic paragangliomas is
particularly suggestive of a mutation in SDHB [9]. The respective estimated risks of
developing pheochromocytoma or paraganglioma range from ~ 1% for patients with
neurofibromatosis to ~ 50% for for those with MEN2, and risks of the tumors becoming
malignant range from <3% for patients with MEN2 to 80% for those with mutated SDHB
[6].
While the above mutations predispose to the development of paragangliomas,
secondary genetic alterations are probably necessary for tumors to develop.
Chromosomal losses involving 1p, 3p, 3q and 11q are frequent in both sporadic and
familial pheochromocytomas or paragangliomas. The rate at which these secondary
changes accrue may account for the variable risk of developing pheochromocytomas or
paragangliomas in hereditary tumor syndromes and the nature of the changes may
contribute to differences in tumor phenotype. The most prevalent abnormality identified
to date is a deletion of a portion of chromosome 1p, also common in neuroblastomas,
containing an unidentified tumor suppressor gene [2]. Although the putative suppressor
gene locus is close to SDHB, the latter does not appear to be the target [12]. Similarly, the
VHL locus is close to the deleted region of 3p. Specific patterns of secondary genetic
changes segregate with specific syndromic mutations, suggesting multiple pathways of
tumorigenesis as recently reviewed by Dannenberg et al [13]. For example, 1p deletions
occur in approximately 80% of MEN2 or sporadic tumors [14]. However, they only are
found in 15% of VHL tumors, which more frequently show losses on chromosome 11
[15], as also reported in sporadic parasympathetic paragangliomas [16].
Gene Expression Profiling
3
Microarray-based gene expression profiling studies complemented by
immunohistochemical and/or biochemical analyses have revealed sets of markers that
tend to be clustered in tumors with specific genetic backgrounds, in subsets of sporadic
tumors, and in benign versus malignant tumors. Two studies now show that
pheochromocytomas/paragangliomas with VHL, SDHB or SDHD mutations form a
cluster with a “transcription signature” characterized by genes associated with hypoxiadriven transcription pathways. In contrast, the signature of tumors with with RET or NF1
mutations includes features consistent with incresed activity of the Ras-mediated MAPK
pathway [6,17]. An additional distinctive characteristic of VHL tumors is that they
usually do not express phenylethanolamine N-methyltransferase, the enzyme that
synthesizes epinephrine from norepinephrine, and are therefore noradrenergic. In
contrast, MEN2 and NF1 tumors produce both epinephrine and norepinephrine [18]. As
with other genotype-phenotype correlations, care must be taken in these analyses to
account for characteristics of the anatomic site of origin of the tumor in question, for
example, that epinephrine production is normally confined to the adrenal medulla. In the
case of VHL, even intra-adrenal tumors have a noradrenergic phenotype. They are also
reported to often have a distinct histological appearance suggestive of hypoxic signaling,
with a thick vascular capsule and many small blood vessels interspersed among tumor
cells [19]
Cell Biology:
The meanings of differences
Hypothetically, several models might be invoked to explain the genotypephenotype correlations that characterize pheochromocytomas/paragangliomas:
Cell of origin
Adrenal chromaffin cells of most species have separate populations of
epinephrine (E) and norepinephrine (NE) cells that can be readily identified by
immunohistochemical staininng for PNMT. An attractively simple model would suggest
for example that pheochromocytomas in VHL disease or MEN2 arise from such
populations. Most (though not all) chromaffin cells in the human adrenal appear to have
a mixed phenotype, making that explanation unlikely. Nonetheless, more subtle
expressions of cellular heterogeneity may be involved. In the adult human adrenal subsets
of medullary cells differentially express various neuropeptides [20] or the RET protooncogene [21]. Recent developmental studies of mouse models suggest that
sympathoadrenal progenitors are to some extent heterogeneous prior to entering the
adrenal primordium and that cell fate commitment within the developing adrenal might
involve several different signaling mechanisms [22-25].
Pathway dependence
A model based entirely on pathway dependence would suggest that signal
transduction pathways activated by a tumorigenic mutation, e.g., a RET mutation in MEN
2, determine both the distribution of tumors and characteristics of those tumors. Although
pathway dependence undoubtedly is involved in genotype-phenotype differences,
4
determining the precise roles of different pathways either in tumorigenesis or phenotype
determination is difficult.
Considerable evidence demonstrates that major signal transduction pathways
suuch as the MAPK cascade can function as “rheostats” rather than “on/off switches”
[26]. Signaling effectors can exert qualitatively as well as quantitatively different effects
depending on levels of pathway activation. Those levels are determined by the amount of
incoming signal (e.g., the concentration of a growth factor [27]), the amount of effector
expressed in a target cell (e.g., growth factor receptor), the intracellular distribution of the
effector (e.g., cytoplasm or membrane lipid rafts [28] and the expression of cooperative
or counteracting signal transducers. Different experimental conditions or models can
therefore lead to different conclusions. In cultures of normal rat chromaffin cells, nerve
growth factor (NGF) can stimulate both proliferation and terminal neuronal
differentiation, both requiring MAPK activation, but the former occuring at 1/10th the
NGF concentration of the latter [29]. Expression of constitutively active MEN2 mutant
RET is mitogenic in fibroblast cultures but is anti-mitogenic and induces neuronal
differentiation in a seemingly more appropriate model of rat pheochromocytoma cells
[30]. Physiological activation of wild type ret is by its ligand, glial cell line derived
neurotrophic factor (GDNF), also induces neuronal differentiation of mouse [31] and
human [32] pheochromocytoma cells.
Functional and anatomic context
Although adrenal chromaffin cells and the chief cells of other sympathetic and
parasympathetic paraganglia are closely related and there is some functional overlap [33],
their utilization in different anatomic contexts subjects them to different types of stimuli.
For example, the adult adrenal medulla responds principally to signals derived from preganglionic sympathetic neurons via trans-synaptic stimulation, while extra-adrenal
sympathetic paraganglia are sparsely innervated and presumably respond to local or
humoral messengers. The chief cells of the carotid body form reciprocal synapses with
glossopharyngeal nerve endings and function in chemosensory reflexes including
oxygen-sensing. Even if cells in all three locations were intrinsically identical, their
exposure to potential tumor promoter effects of neurotransmitters or other secretogogues
would be different, so that a potentially oncogenic mutation expressed in all cells would
be manifest by preferential development of tumors in different locations.
In fact, all three of the above probably contribute to genotype-phenotype correlations,
although their precise contributions are almost entirely unknown. Intrinsic cell- or tissuespecific differences in expression of critical signal transducers could confer susceptibility
to anatomically dependent signals. Nerve endings containing different neurotransmitters
[34] could confer functional heterogeneity within cell populations that appear
homogeneous.
A further overlay that must be considered is phenotype plasticity. Recent studies of the
rat adrenal medulla suggest that neurally derived signals may increase the expression of
receptors that regulate chromaffin cell function, including the receptor tyrosine kinase
RET. The finding that RET expression is not static may help to resolve the conundrum of
how that molecule, which is expressed at very low levels in the adult adrenal, contributes
5
to the development of adrenal medullary hyperplasia and pheochromocytoma in adults
with MEN2 syndromes [35].
The search for similarities: Cross-talk, common denominators and unifying hypotheses
Signaling pathways, receptors and ion channels involved in the function and
development of the adrenal medulla and related tissues cross communicate at multiple
levels. For example RET encodes a developmentally regulated receptor tyrosine kinase
involved in formation and maintenance of the nervous system. Gain -of -function RET
mutations in MEN 2A cause constitutive kinase activation, one consequence of which is
activation of Ras signaling. NF1 encodes a tumor suppressor gene, the functions of
which include Ras GTPase activity that may terminate Ras signaling. Loss-of-function
NF1 mutations in neurofibromatosis might therefore increase Ras signaling as in
MEN2A. VHL encodes a protein that guides proteolytic degradation of hypoxiainducible transcription factors (HIFs) and other proteins marked for destruction by
oxygen-sensitive prolyl hydroxylases, and loss-of-function VHL mutations mimic
hypoxia by interfering with this oxygen-sensitive proteolysis. SDH enzymes are
components of mitochondrial complex 2 and function both in the Krebs cycle, converting
succinate to fumarate, and as components of the electron transport chain. Loss-offunction mutations of SDH hypothetically could mimic hypoxic signaling both through
accumulation of succinate and from buildup of reactive oxygen species. However,
succinate accumulation may be more important in that oxygen-sensitive prolyl
hydroxylases that mark proteins for VHL-guided degradation utilize alpha-ketoglutatate
(2-oxoglutarate), the Krebs cycle precursor of succinate, as a co-substrate, generating
succinate as a by-product. Excess succinate may cause competitive inhibition of those
hydroxylases, thereby indirectly affecting the functions of VHL [36,37]. Hypoxia can
also activate p42/p44 MAPK, which are downstream effectors of Ras signaling, through
several pathways [33,38]. These effects could occur both in sympatheitic and
parasympathetic paragangliomas, as shown by the fact that experimental
pheochromocytoma cells express the same oxygen-sensing mechanisms as carotid body
chief cells [33].
The concept of cross talk provides a general framework for understanding how
mutations of apparently unrelated genes might lead to the same type of tumor by
convergent signaling. In addition, they suggest that targeted therapies might be directed
either at a specific mutated gene or at a downstream signal transducer. However, it
remains to be determined what precise mechanisms lead to tumor formation. In an
intriguing recent paper, Lee et al propose that the mutations of RET, NF1, VHL and SDH
in hereditary pheochromocytoma/paraganglioma predispose to tumor formation by
causing defective developmental culling, promoting survival of damaged cells that would
normally destroyed by apoptosis during embryogenesis [39]. An important argument in
favor of survival rather than mitogenesis as the major common denominator of the
syndrome-associated genes is the rarity of the same mutations in sporadic
pheochromocytomas/paragangliomas, suggesting that the mutations only need to act
during a limited developmental window. The proposed model was based on studies of
6
PC12 rat pheochromocytoma cells and the specific convergence point was a prolyl
hydroxylase known as EglN3, thus far not found mutated or overexpressed in human
pheochromocytomas/paragangliomas. Nevetheless, other genes might serve the same
purpose. An important implication of the model is that tumor precursor cells could
theoretically be identified and eradicated in the adrenal glands or paraganglia of
individuals who carry RET, NF1, VHL or SDH mutations.
Interestingly, the survival promoting effect of mutant VHL in the model proposed
by Lee et al does not appear to be predominantly HIF-dependent [39]. This model is
thereby the first to adequately account for the fact that VHL2C mutations do not prevent
HIF degradation, serving as a reminder that tumor suppressor genes should not forever be
conceptually pigeonholed according to the functions intitially ascribed to them as was the
case for VHL and HIF. A similar caveat might be applied to NF1 in that it is one of the
largest genes in the entire genome and has mutations distributed throughout its length
rather than in discrete hotspots as for RET, VHL or SDH. Most NF1 mutations are distant
from the Ras-GTPase domain, and evidence of abnormal Ras signaling has not been
detected in pheochromocytoma cell lines from nf1 knockout mice [40].
References
1.
2.
3.
4.
5.
6.
7.
8.
9.
DeLellis RA, Lloyd RV, Heitz PU, Eng C: Tumours of Endocrine Organs. In
World Health Organization Classification of Tumors. Lyon, IARC Press, 2004
Linnoila RI, Keiser HR, Steinberg SM, Lack EE: Histopathology of benign versus
malignant sympathoadrenal paragangliomas: clinicopathologic study of 120 cases
including unusual histologic features. Hum Pathol 21:1168-80, 1990
Lloyd RV, Sisson JC, Shapiro B, Verhofstad AA: Immunohistochemical
localization of epinephrine, norepinephrine, catecholamine-synthesizing enzymes,
and chromogranin in neuroendocrine cells and tumors. Am J Pathol 125:45-54,
1986
Bryant J, Farmer J, Kessler LJ, Townsend RR, Nathanson KL:
Pheochromocytoma: the expanding genetic differential diagnosis. J Natl Cancer
Inst 95:1196-204, 2003
Neumann HP, Bausch B, McWhinney SR, et al.: Germ-line mutations in
nonsyndromic pheochromocytoma. N Engl J Med 346:1459-66., 2002
Eisenhofer G, Bornstein SR, Brouwers FM, et al.: Malignant pheochromocytoma:
current status and initiatives for future progress. Endocr Relat Cancer 11:423-36,
2004
Baysal BE: Genomic imprinting and environment in hereditary paraganglioma.
Am J Med Genet C Semin Med Genet 129:85-90, 2004
Schiavi F, Boedeker CC, Bausch B, et al.: Predictors and prevalence of
paraganglioma syndrome associated with mutations of the SDHC gene. Jama
294:2057-63, 2005
Astuti D, Hart-Holden N, Latif F, et al.: Genetic analysis of mitochondrial
complex II subunits SDHD, SDHB and SDHC in paraganglioma and
phaeochromocytoma susceptibility. Clin Endocrinol (Oxf) 59:728-33, 2003
7
10.
11.
12.
13.
14.
15.
16.
17.
18.
19.
20.
21.
22.
23.
24.
25.
26.
Braun S, Riemann K, Kupka S, et al.: Active succinate dehydrogenase (SDH) and
lack of SDHD mutations in sporadic paragangliomas. Anticancer Res 25:2809-14,
2005
Dahia PL, Hao K, Rogus J, et al.: Novel pheochromocytoma susceptibility loci
identified by integrative genomics. Cancer Res 65:9651-8, 2005
Astuti D, Morris M, Krona C, et al.: Investigation of the role of SDHB
inactivation in sporadic phaeochromocytoma and neuroblastoma. Br J Cancer
91:1835-41, 2004
Dannenberg H, Komminoth P, Dinjens WN, Speel EJ, de Krijger RR: Molecular
genetic alterations in adrenal and extra-adrenal pheochromocytomas and
paragangliomas. Endocr Pathol 14:329-50, 2003
Bender BU, Gutsche M, Glasker S, et al.: Differential genetic alterations in von
Hippel-Lindau syndrome- associated and sporadic pheochromocytomas. J Clin
Endocrinol Metab 85:4568-74., 2000
Lui WO, Chen J, Glasker S, et al.: Selective loss of chromosome 11 in
pheochromocytomas associated with the VHL syndrome. Oncogene 21:1117-22,
2002
Bikhazi PH, Messina L, Mhatre AN, Goldstein JA, Lalwani AK: Molecular
pathogenesis in sporadic head and neck paraganglioma. Laryngoscope 110:13468, 2000
Dahia PL, Ross KN, Wright ME, et al.: A HIF1alpha Regulatory Loop Links
Hypoxia and Mitochondrial Signals in Pheochromocytomas. PLoS Genet 1:e8,
2005
Eisenhofer G, Goldstein DS, Kopin IJ, Crout JR: Pheochromocytoma: rediscovery
as a catecholamine-metabolizing tumor. Endocr Pathol 14:193-212, 2003
Koch CA, Mauro D, Walther MM, et al.: Pheochromocytoma in von hippellindau disease: distinct histopathologic phenotype compared to
pheochromocytoma in multiple endocrine neoplasia type 2. Endocr Pathol 13:1727, 2002
Lundberg JM, Hamberger B, Schultzberg M, et al.: Enkephalin- and somatostatinlike immunoreactivities in human adrenal medulla and pheochromocytoma. Proc
Natl Acad Sci U S A 76:4079-83, 1979
Powers JF, Brachold JM, Tischler AS: Ret protein expression in adrenal
medullary hyperplasia and pheochromocytoma. Endocr Pathol 14:351-61, 2003
Gut P, Huber K, Lohr J, et al.: Lack of an adrenal cortex in Sf1 mutant mice is
compatible with the generation and differentiation of chromaffin cells.
Development 132:4611-9, 2005
Unsicker K, Huber K, Schutz G, Kalcheim C: The chromaffin cell and its
development. Neurochem Res 30:921-5, 2005
Huber K, Karch N, Ernsberger U, Goridis C, Unsicker K: The role of Phox2B in
chromaffin cell development. Dev Biol 279:501-8, 2005
Ernsberger U, Esposito L, Partimo S, et al.: Expression of neuronal markers
suggests heterogeneity of chick sympathoadrenal cells prior to invasion of the
adrenal anlagen. Cell Tissue Res 319:1-13, 2005
Hazzalin CA, Mahadevan LC: MAPK-regulated transcription: a continuously
variable gene switch? Nat Rev Mol Cell Biol 3:30-40, 2002
8
27.
28.
29.
30.
31.
32.
33.
34.
35.
36.
37.
38.
39.
40.
Ma Y, Campenot RB, Miller FD: Concentration-dependent regulation of neuronal
gene expression by nerve growth factor. J Cell Biol 117:135-41, 1992
Encinas M, Tansey MG, Tsui-Pierchala BA, Comella JX, Milbrandt J, Johnson
EM, Jr.: c-Src is required for glial cell line-derived neurotrophic factor (GDNF)
family ligand-mediated neuronal survival via a phosphatidylinositol-3 kinase (PI3K)-dependent pathway. J Neurosci 21:1464-72, 2001
Powers JF, Shahsavari M, Tsokas P, Tischler AS: Nerve growth factor receptor
signaling in proliferation of normal adult rat chromaffin cells. Cell Tissue Res
295:21-32, 1999
Rossel M, Pasini A, Chappuis S, et al.: Distinct biological properties of two RET
isoforms activated by MEN 2A and MEN 2B mutations. Oncogene 14:265-75,
1997
Powers JF, Schelling K, Brachold JM, et al.: High-level expression of receptor
tyrosine kinase ret and responsiveness to ret-activating ligands in
pheochromocytoma cell lines from neurofibromatosis knockout mice. Mol Cell
Neurosci 20:382-9., 2002
Powers JF, Tsokas P, Tischler AS: The ret-activating ligand GDNF is
differentiative and not mitogenic for normal and neoplastic human chromaffin
cells in vitro. Endocrine Pathol 9:325-331, 1998
Spicer Z, Millhorn DE: Oxygen sensing in neuroendocrine cells and other cell
types: pheochromocytoma (PC12) cells as an experimental model. Endocr Pathol
14:277-91, 2003
Holgert H, Holmberg K, Hannibal J, et al.: PACAP in the adrenal gland-relationship with choline acetyltransferase, enkephalin and chromaffin cells and
effects of immunological sympathectomy. Neuroreport 8:297-301, 1996
Powers JF, Brachold JM, Ehsani SA, Tischler AS: Up-regulation of ret by
reserpine in the adult rat adrenal medulla. Neuroscience 132:605-12, 2005
Schofield CJ, Ratcliffe PJ: Signalling hypoxia by HIF hydroxylases. Biochem
Biophys Res Commun 338:617-26, 2005
Hausinger RP: FeII/alpha-ketoglutarate-dependent hydroxylases and related
enzymes. Crit Rev Biochem Mol Biol 39:21-68, 2004
Conrad PW, Freeman TL, Beitner-Johnson D, Millhorn DE: EPAS1 transactivation during hypoxia requires p42/p44 MAPK. J Biol Chem 274:33709-13,
1999
Lee S, Nakamura E, Yang H, et al.: Neuronal apoptosis linked to EglN3 prolyl
hydroxylase and familial pheochromocytoma genes: developmental culling and
cancer. Cancer Cell 8:155-67, 2005
Park JI, Powers JF, Tischler AS, Strock CJ, Ball DW, Nelkin BD: GDNF-induced
leukemia inhibitory factor can mediate differentiation via the MEK/ERK pathway
in pheochromocytoma cells derived from nf1-heterozygous knockout mice. Exp
Cell Res 303:79-88, 2005
9