Research
Overview
People
Profiles
How to Reach Us
Contact Info
Research Overview
Research in my lab focuses on five important areas at the interface between molecular cell biology and medicine:
A. Erythropoietin receptor (EpoR) and red cell development.
Red blood cell development, especially on regulation of terminal proliferation and differentiation of erythroid progenitor cells by transcription factors activated by signals downstream of the erythropoietin receptor and cell surface integrins, and on the mechanisms of chromatin condensation and enucleation;
Hematopoietic stem cells, identifying the stromal cells in the fetal liver and other organs that support stem cell self- renewal in vivo, and identifying novel growth factors that support their expansion in culture;
C. MicroRNAs that modulate differentiation.
MicroRNAs, defining their roles in lineage commitment of hematopoietic stem and progenitor cells, and regulating adipocyte and muscle differentiation and function;
D. Hormones controlling fatty acid and glucose metabolism.
Adiponectin, a hormone we cloned that is made exclusively by fat cells and that increases fatty acid and glucose metabolism by muscle, and several adiponectin orthologs that share adiponectin’s biological activities;
E. New signaling pathways and new technologies.
Regulated cleavage and release of the extracellular domain ("ectodomain shedding") of transmembrane precursors of several secreted growth factors.
What ties all of these projects together is their focus on the basic cell and molecular biology of proteins important for human physiology and disease. Another is the development and use of powerful tools of molecular biology including expression cloning strategies such as those we used to clone most of the receptors, transporters, and several signaling proteins we study, techniques for immunolocalizing proteins in cells and tissues, and generation and analysis of many types of gene- altered mice.
Erythropoietin (Epo) controls production of red blood cells; it is produced by the kidney in response to low oxygen pressure in the blood. Epo binds to Epo receptors on the surface of committed erythroid late BFU-E and CFU-E progenitors, blocking apoptosis (programmed cell death), their usual fate, and triggering them to undergo a two- day program of 4 – 5 terminal erythroid cell divisions and differentiation. Following condensation of chromatin and subsequent enucleation reticulocytes (immature red cells) are released into the blood.
As evidenced by the properties of Epo- and EpoR- deficient mice we generated, Epo and the EpoR are essential for proliferation and differentiation of committed erythroid progenitors, as is the cytosolic protein-tyrosine kinase JAK-2. JAK2 binds to the EpoR cytosolic domain in the endoplasmic reticulum and facilitates its folding to promote cell surface expression. EpoRs exist as inactive dimers on the cell surface; Epo binding changes their conformation, leading to JAK2 transphosphorylation and activation.
JAK2 activates many signaling proteins including the transcription factor
Stat5, the PI-3’ kinase Akt kinase pathway, and the Ras MAPK pathway.
These pathways interact to prevent apoptosis of committed erythroid progenitors
allowing them to undergo a predetermined program of terminal proliferation
and erythroid differentiation. We showed that Stat5 directly activates
transcription of the anti-apoptotic protein bclxL. Stat5-/-
mice exhibit fetal anemia and increased apoptosis of erythroid progenitors
caused by reduced bclxL levels. Adult Stat5-/- mice
are anemic and deficient in generating high erythropoietic rates in response
to stress. Thus Stat5 controls one rate-determining step regulating early
erythroblast survival.
Activation of the PI-3’ kinase pathway leads to activation of the Akt kinase and then phosphorylation and inhibition of FOXO3a, a member of the Forkhead transcription factor family. FOXO3a, in turn, activates transcription of Tumor Necrosis Factor Apoptosis-Inducing Ligand (TRAIL). We showed that inhibition of TRAIL production by Epo addition partially rescues cells from apoptosis, demonstrating the importance of this pathway in red cell formation. Additionally, we showed that activated Akt phosphorylates the erythroid important transcription factor GATA-1 both in vitro and in vivo and enhances GATA-1 activity in erythroid cells.
Activation of Epo receptor signaling How Epo stimulation
activates EpoR-associated JAK2 is a central question in cytokine receptor
signaling. As a monomeric, asymmetric molecule, Epo employs two different
interfaces, termed site 1 and site 2, to bind to the two monomers of an
EpoR homodimer. Because the EpoR dimer is comprised of two identical transmembrane
proteins, it has been impossible to determine whether the EpoR bound to
Site 1 signals identically or differently from the one bound to Site 2.
Mala L. Radhakrishnan, a student in Professor Bruce Tidor’s lab
in the MIT Department of Bioengineering, computationally identified amino
acids in the EpoR important for binding to Site 1 in Epo but not site
2, and vice versa. She designed, and Lucy Zhang and Xiaohui Liu experimentally
created and verified, two mutant Epo receptors - one that binds only to
Epo site 1 but not site 2, and the other to site2 and not site 1. Expression
of either mutant receptor alone in Ba/F3 cells cannot elicit a signal
in response to erythropoietin, but when co-expressed in the same cell
there was a proliferative response and activation of the JAK2 Stat5 signaling
pathway. A truncated erythropoietin receptor with only one cytosolic tyrosine
(Y343; important for activation of Stat5) on only one receptor monomer
is sufficient for signaling in response to erythropoietin, regardless
of the monomer on which it is located.
Several years ago we identified three conserved hydrophobic residues in
the juxtamembrane cytosolic domain of EpoR, L253, I257, and W258, and
showed these are necessary for Epo- triggered activation of the associated
JAK2. Mutating any of these residues to alanine in the EpoR homodimer
dramatically suppressed growth and signaling responses to Epo stimulation
but did not affect the ability of the Epo receptors to bind Jak2, traffic
normally to the cell surface, or bind Epo. Using our Site-1 deficient
and Site-2 deficient EpoRs we showed that these conserved residues could
be located either on the Site 1- or the Site 2- binding receptor. We concluded
that despite asymmetry in the ligand-receptor dimer interaction, both
sides are competent for signaling, and we suggest that the receptors signal
equally.
Mutations that activate JAK2 Disregulation of JAK2 signaling
has been implicated in several hematological malignancies. One example,
an acquired point mutation in the JAK2, V617F, was recently discovered
in most patients with polycythemia vera and half of those with other myeloproliferative
disorders. It is unclear why this oncogenic JAK2 mutation is reponsible
for such related yet very distinct diseases. In collaboration with Dr.
Gary Gilliland’s lab in Harvard Medical School, Xiaohui Liu and
Wei Tong showed that both cell transformation by JAK2V617F and constitutive
activation of the JAK/STAT signaling pathway requires the presence of
cognate type I cytokine receptors. Using IL-3-dependent Ba/F3 cells and
32D cells, Xiaohui and Wei showed that expression of JAK2V617F confers
factor-independent growth only in cells coexpressing homodimeric cytokine
receptors such as the erythropoietin receptor, thrombopoietin receptor,
or granulocyte colony-stimulating factor receptor. Furthermore, these
cytokine receptors are also required for the constitutive phosphorylation
and activation of both JAK2 and STAT5. By coexpressing JAK2V617F together
with EpoR mutants, they demonstrated that EpoR provides an essential scaffold
for factor-independent activation of JAK2V617F and STAT5; in particular
phosphorylation of tyrosines in the cytosolic domain of the EpoR is essential
for activation of STAT5. We concluded that JAK2V617F transduces oncogenic
signals in conjunction with cytokine receptors, in a cytokine-independent
version of its normal signaling mechanisms. These findings provide a molecular
basis for the prevalence of JAK2V617F in diseases of myeloid cells that
express Type I cytokine receptors, and explain the overlapping clinical
observations in these diseases.
Wei Tong extended these results in collaboration with Dr. Gary Gilliland’s
lab and that of Dr. Tony Green of the University of Cambridge. While the
JAK2 V617F mutation is found in many patients with myeloproliferative
disorders the molecular basis for most V617F-negative patients with a
myeloproliferative disorder was unclear. In patients diagnosed with V617F-negative
polycythemia vera or idiopathic erythrocytosis four different somatic
gain-of-function mutations affecting JAK2 exon 12 were identified. Patients
with a JAK2 exon 12 mutation presented with a histologically distinct
variant of polycythemia vera, characterized by an isolated erythrocytosis,
erythropoietin-independent erythroid colonies, and suppressed serum erythropoietin.
In contrast to the V617F mutation, exon 12 mutations were associated with
an absence of detectable mutation-homozygous erythroid progenitors. Exon
12 mutations resulted in cytokine-independent proliferation of BaF3 cells
expressing the EpoR cells, together with increased activation of downstream
signaling pathways. Thus JAK2 exon 12 mutations give rise to a distinct
variant of polycythemia vera. Our results emphasize the importance of
molecular genetics for the classification and diagnosis of the myeloproliferative
disorders, and suggest that the different clinical phenotypes associated
with JAK2 exon 12 and V617F mutations may reflect stronger signaling by
the former.
Most researchers agree that cognate cytokine receptors, such as EpoR,
are important for the oncogenic activities of JAK2V617F. However, others
reported contradictory results in Ba/F3 cells, suggesting JAK2V617F is
constitutively active by itself. To this end, Xiaohui demonstrated that
differences in JAK2V617F expression levels likely account for this critical
discrepancy, and overexpressed JAK2V617F functions through binding to
unidentified cytokine receptor dimers. Xiaohui further demonstrated that
receptor-mediated dimerization is critical for the activity of JAK2V617F,
and the EpoR intracellular hydrophobic motif (L253, I257 and W258) that
is essential for Epo-mediated JAK2 activation is not required for the
activation of JAK2V617F. These results support the notion that receptor-mediated
dimerization of JAK2V617F is required and sufficient for its activation.
Furthermore, the constitutively activated JAK2V617F may have subtle structural
differences from the Epo-activated wild-type JAK2, raising the possibility
of developing a JAK2V617F-specific inhibitor.
Trafficking of the Epo receptor In an ongoing collaboration,
Alec Gross used computational models generated in Dr. Bruce Tidor’s
laboratory to better understand how Epo binds to EpoR in different cellular
compartments and how this affects cellular trafficking. Computational
molecular modeling predicted that the EpoR has a naturally occurring “Histidine
Switch” in one of the binding interfaces between Epo and the EpoR.
That is to say, at the neutral pH of the cell surface, this crucial histidine
will be uncharged, but at the more acidic pH in endosomes this histidine
will become protonated and a positive charge will be introduced into the
binding interface. The prediction was that the presence of a positive
charge at this position will decrease the affinity of Epo for EpoR, and
thus this histidine will play an important role in dissociating the Epo-
EpoR complex and in the cellular trafficking of Epo•EpoR complexes
following their internalization.
Alec experimentally tested these predictions and showed that there is
indeed a naturally occurring Histidine Switch in the EpoR. Binding of
Epo to EpoR is strongly pH-dependent, as shown by a greatly decreased
affinity of Epo for EpoR at pH 5.5 compared to pH 7.3. When Alec mutated
the histidine in question to amino acids with non-pH- sensitive side chains,
the affinity of Epo for EpoR was only slightly decreased at pH 5.5 compared
to pH 7.3. Thus, as predicted by molecular modeling, a single histidine
in the EpoR is primarily responsible for the pH-dependent binding of Epo.
Trafficking of Epo was altered in cells expressing the mutant EpoRs that
have decreased pH-dependent binding: intact Epo was retained intracellularly
for an abnormally extended time and Epo degradation was delayed. These
results suggested that sorting steps in early recycling endosomes, where
the pH is only mildly acidic, are not affected by pH-dependent binding
of Epo to EpoR. However, pH-dependent binding of Epo plays an important
role in later sorting and degradation steps. Under this scenario, if Epo
does not dissociate from EpoR in the more acidic late endosomes, as with
the histidine-mutant EpoRs, there is a delay in trafficking to and/or
the actual degradation of Epo in lysosomes.
Transcriptional control of gene expression during erythroid differentiation. Shilpa Hattangadi’s project involves determining the transcriptional regulatory networks governing the important changes in gene expression that occur during terminal proliferation and differentiation of erythroid precursors. She began with expression profiling during erythroid differentiation: by isolation of mRNA from purified erythroid precursors in successive differentiation stages followed by hybridization to DNA microarrays and eventual confirmation of select gene’s expression by qRT-PCR. Initial expression profiling results indicate that major changes in gene regulation occur during erythroblast differentiation, comporaneously with induction of Ter119 expression, an erythroid- specific surface protein, and hemoglobin. Upregulated genes include those involved in hemoglobin metabolism, heme and porphyrin ring metabolism, cell and nuclear membrane structure, iron ion homeostasis, negative regulators of cell cycle, oxygen transport, and metabolism of oxygen and reactive oxygen species, among others. Genes that were significantly downregulated included genes involved in TNF-alpha production, NADP metabolism, NF-kappaB binding, actin binding, ubiquitin protein ligation, and non-erythroid specific functions such as immune response, antigen stimulation and response, and phagocytosis. Further cluster analysis is ongoing, revealing interesting sets of genes for further study and future location analysis experiments. She is currently studying the function of a few transcriptional cofactors by shRNA and overexpression experiments in the in vitro erythroid culture model.
The second tier of Shilpa’s project, done in collaboration Bill
Wong, a new postdoctoral fellow, and members of Rick Young's laboratory,
involves immunoprecipitation of chromatin with antibodies specific for
various erythroid- important transcription factors (ChIP), followed by
either hybridization of the recovered DNA to a promoter DNA microarray
(ChIP-chip) or by sequencing of the resulting DNA fragments (ChIP-Seq).
This protocol will enable them to determine all of the genes that have
critical erythroid-important transcription factors bound to their promoter/
enhancer segments. Initial studies focus on transcriptional activation
by Stat5, GATA1, FOG, and Foxo3 but other factors will soon be investigated.
Shilpa's long- term goal is to understand how the complex pattern of gene
expression during terminal erythroid differentiation is regulated by transcription
factors activated initially by signal transduction pathways downstream
of the EpoR, but active in precursors no longer dependent on erythropoietin.
The role of integrins in terminal proliferation and differentiation of purified fetal liver erythroid cells. Fibronectin is an important part of the erythroid niche but its precise role in erythropoiesis is unknown. We showed some years ago that adhesion of these progenitors to fibronectin is essential for normal erythroid development. Shawdee Eshghi showed that both alpha4beta1 and alpha5beta1 integrins are present on erythroid progenitors and support binding of erythroid cells to different fibronectin domains. Shawdee also showed that loss of both alpha4beta1 and alpha5beta1 integrins during erythroid differentiation parallels the loss of adhesion of erythroid cells to fibronectin; this allows retention of immature erythroid cells in the bone marrow and then release of the enucleated reticulocytes into the circulation.
Using our in vitro culture system Shawdee showed that alpha4beta1 integrin is required for terminal proliferation of erythropoietic progenitors. More specifically, alpha4beta1 integrin and erythropoietin mediate temporally distinct steps in erythropoiesis. By culturing fetal liver erythroid progenitors she showed, in a collaboration with the lab of Prof. Richard Hynes, that fibronectin and Epo regulate erythroid proliferation in temporally distinct steps: an early Epo-dependent phase is followed by a fibronectin-dependent phase. In each phase, Epo and fibronectin promote expansion by preventing apoptosis, in part through enhancing expression of the antiapoptotic protein bcl-xL. By culturing erythroid progenitors on recombinant fibronectin fragments she established that only substrates that engage alpha4beta1 integrin support normal proliferation. Taken together, these data suggest a two-phase model for growth factor and extracellular matrix regulation of erythropoiesis, with an early Epo-dependent, integrin-independent phase followed by an Epo-independent, alpha4beta1 integrin-dependent phase.
Chromatin condensation and enucleation in late stage erythroblasts Mammalian erythroid cells undergo enucleation during a late stage of differentiation, a process that does not occur in other vertebrates. This process has critical physiological and evolutional significance for the morphogenesis and hemoglobin enrichment of mature mammalian red blood cells. Although enucleation has been known for decades the mechanisms that regulate the process remain obscure. Using the new in vitro culture system of fetal liver erythroid progenitors Jing developed, Peng Ji is investigating the mechanism of mammalian erythroid cell enucleation. This and many of our other studies on EpoR signal transduction make use of the system Jing Zhang developed; purified fetal liver erythroid progenitors (so-called CFU-Es) are plated on fibronectin- coated dishes and cultured in the presence of Epo; they undergo normal terminal proliferation and differentiation that can be followed on a single cell level by FACS.
Since actin filaments have been shown to be critical for enucleation,
Peng determined the role of different Rho GTPases, the master regulators
of actin nucleation, on enucleation. Peng showed that deregulation of
Rac GTPase during the late stages of erythropoiesis completely blocks
enucleation of cultured mouse fetal erythroblasts without affecting their
normal proliferation and differentiation. The contractile actin ring formed
on the plasma membrane of late-stage erythroblasts at the boundary between
the cytoplasm and nucleus of enucleating cells was disrupted when Rac
GTPase was inhibited in late stages of erythropoiesis. Peng further demonstrated
that the Rac GTPase activity is mediated by a downstream target protein,
mDia2, a formin protein required for nucleation of unbranched actin filaments.
These results reveal important roles for Rac GTPase and mDia2 in enucleation
of mammalian erythroblasts.
Peng, together with Francisco Sanchez-Rivera, is also focusing on the role of histone deacetylases (HDACs) in chromatin and nuclear condensation and enucleation of late erythroid cells. They showed that inhibition of HDAC activities by Trichostatin A completely blocked enucleation, and that specific inhibition of HDAC6 activity partially blocked enucleation. Peng further showed that mDia2 is acetylated in vivo and his current aim is to determine whether HDAC6 can deacetylate mDia2 and in so doing promote red cell enucleation. In parallel they are investigating the roles of HDACs in inactivating gene transcription and condensing chromatin prior to enucleation. In collaboration with Tzutzuy Ramirez, a fellow with Dr. Maki Murata Hori of the Temasek Life Sciences Laboratory, Singapore, they are investigating the roles of many cytoskeletal and other proteins in nuclear migration and enucleation of these cells, in part using video microscopy of cells expressing fluorescent- tagged proteins.
At the same time, Peng is also interested in the roles of mDia2 on hematopoietic
stem cell homing and migration. He is currently generating an mDia2 knockout
mouse model and he will use these mice to study the roles of mDia2 in
hematopoiesis.
Diamond Blackfan anemia and other ribosomal protein diseases Diamond Blackfan anemia (DBA) is a congenital anemia and broad developmental disease that develops at birth or soon after. The anemia is due to failure of production of erythrocytes and their precursors, with normal or near normal myeloid cells and platelets. DBA is inherited in about 10-20% of cases, mostly as an autosomal dominant. Genetic studies have led to the surprising identification of mutations in a ribosomal protein (RP) gene, RPS19, in 20 - 25% of both familial and sporadic cases. Recently Colin Sieff and colleagues discovered a mutation in another ribosomal protein gene, RPS24 that co-segregates with affected family members in a large DBA pedigree, and has also identified mutations in three large subunit RPs. While experiments in yeast and recently in mammalian cells show that RPS19 depletion or mutation leads to a block in ribosomal RNA biosynthesis, this result does not explain why erythropoiesis is so severely affected in DBA.
One hypothesis is that during fetal development immature erythroid cells
proliferate more rapidly than other lineages and therefore require very
high ribosome synthetic rates to generate sufficient capacity for translation
of erythroid specific transcripts that must take place before these unique
cells enucleate. To test this hypothesis Colin, assisted by Jing Yang,
first purified the most primitive erythroid progenitor cells. During the
first 24 hours culture in the presence of Epo cell number increases 3-4
fold; remarkably, there is a 6-fold increase in RNA content during the
same period, suggesting that the cells accumulate an excess of ribosomal
RNA and ribosomes during early erythropoiesis. siRNA RPS19 knockdown cells
show reduced proliferation at 48 hours of culture, but the differentiation
pattern of the surviving knockdown positive cells is similar to that of
the controls.
While RPS19 mRNA is rapidly depleted by the siRNA, Western analysis does not show a deficiency of RPS19 protein. This suggests strongly that the proliferative defect is not due to insufficiency of RPS19 protein, and is more likely due to a block in ribosome biogenesis that leads to nucleolar stress. Preliminary data show that apoptosis is increased, cells are blocked in the G1 stage of the cell cycle, and p53 is increased. Nucleolar “stress” could lead to redistribution of cell cycle proteins normally resident in the nucleolus with consequent p53 mediated cell cycle arrest and/ or apoptosis. Alternatively, increased levels of blocked rRNA processing intermediates and/or ribosome assembly intermediates could lead to defects in chromatin condensation and/or mitosis - possibly due to effects on nucleolar condensation. These effects, which have been demonstrated in other cell lineages, would be exacerbated in erythroid cells because of the amazingly quick cell cycle. In summary, cell cycle deficiencies can explain the erythroid deficiency in DBA.
How glucocorticoids stimulate erythropoiesis The therapeutic regimen for DBA consists of blood transfusions and glucocorticoid (GC) treatment. Interestingly in 40% of DBA patients GCs are able to rescue the erythoid progenitor defect. Although the mechanism for this is not known, it demonstrates an important role for GCs in red cell development, possibly enhancing the effects of Epo. Other clinical examples of how GCs affect red cell development are Morbus Addison disease (low cortisol levels), which is associated with severe anemia, and Cushing’s syndrome (high cortisol levels), which curiously often is associated with increased red blood cell counts. Johan Flygare is investigating the role of glucocorticoids in red cell development using in vitro culture systems and glucocorticoid receptor (GCR)-deficient mice. Johan’s aim is to first determine at which stages in red cell development GCs are acting and then study these erythroid progenitor populations and try to understand the molecular mechanism for the increased red cell output. Since the main action of the activated GCR is to interact with chromatin and regulate transcription Johan hopes to answer many questions by mapping exactly where in the chromatin the activated receptor binds by ChIP-Seq, which binding partners it has and how transcription is repressed and/or activated at these sites.
A cell culture assay for gene- damaging chemicals Joe Shuga, in collaboration with the laboratories of Profs Leona Samson and Linda Griffith, extended our in vitro culture system for erythroid progenitors into an assay for genotoxicity. Assays that predict toxicity are an essential part of drug development and many drugs fail in phase I clinical trials; therefore, there is a demand for models that can better predict human responses. The mouse in vivo micronucleus assay is a robust toxicity test that assesses the genotoxic effect of drugs- especially those that induce breaks in DNA - by detecting chromosome fragments that remain in the reticulocyte after enucleation; an in vitro correlate to this assay might allow extension to human cells and thus better predictive power in drug development. As first steps in developing a toxicity assay, Joe has adapted our in vitro erythropoiesis culture system to induce optimized erythropoietic growth from the Lin- population of adult murine bone marrow. Using this system he demonstrated that exposure to genotoxicants induces micronucleus -formation in this culture system. In particular, Joe showed that addition of the alkylating agents BCNU, MNNG, and MMS to this culture system induces both a cytotoxic response and an increase in micronucleus frequency within the reticulocyte population. This increase in micronucleus production following exposure to these alkylating genotoxicants provides a clear signal of the genotoxic mechanism that likely induced the observed erythropoietic toxicity.
Hematopoietic stem cells (HSCs) are defined by their ability to self-renew and to differentiate into all blood cell types. These very rare cells – about 1:10,000 to 1:50,00 cells in bone marrow - form the basis of bone marrow transplantation for treatment of leukemia and other cancers, and are also a promising cell target for developing gene therapies for treating a broad variety of human diseases. However, development of these important clinical applications of HSCs are greatly hampered by the lack of understanding of the extracellular and intracellular signals that govern their fates and the difficulty in ex vivo expansion of these cells. We quantitate these cells by bone marrow transplantation, monitoring the long- term repopulation of the hematopoietic compartment of lethally irradiated mice. This assay requires several months to complete.
Novel growth factors for hematopoietic stem cells No
single known growth factor or combination of growth factors reproducibly
supported HSC expansion in culture. Furthermore existing lines of “supportive
stromal cells” did not support expansion of HSCs; at best they maintained
the level of HSCs over time, presumably due to a steady state between
generation of new HSCs by division and differentiation of “old”
stem cells. Thus Chengcheng Zhang, assisted by Megan Kaba, turned to mouse
fetal liver since the number of fetal HSCs normally increased markedly
between embryonic Day 14 and Day 21. Chengcheng hypothesized that unknown
growth proteins are produced by as- yet unidentified populations of fetal
liver cells that stimulate the expansion of fetal liver HSCs. He then
identified Embryonic Day 15 fetal liver CD3+ Ter119- cells as a novel
cell population that supports a net expansion of HSC numbers in culture.
CD3 is generally thought to be a specific T-cell marker; By transcriptional
profiling of these cells and several others that do not support HSC expansion,
Chengcheng uncovered several novel growth factors that, together, supported
an unprecedented extent of ex vivo expansion of bone marrow HSCs. First
he identified insulin - like growth factor 2 (IGF - 2); a serum- free
culture medium containing low levels of stem cell factor (SCF), thrombopoietin
(TPO), IGF-2, and FGF-1 supported a greater than 8-fold increase in long-term
HSC numbers after a 10-day culture of enriched BM stem cells. Strikingly,
the surface phenotype of ex vivo expanded HSCs was different from that
of freshly isolated HSCs, but this plasticity of surface phenotype did
not significantly alter their repopulation capability.
More recently Chengcheng showed that Angiopoietin-like 2 and 3, secreted proteins specifically produced by day 15 fetal liver CD3+ Ter119- cells, are novel hormones that also stimulate ex vivo expansion of HSCs. Chengcheng showed that, when used in serum- free media in combination with other growth factors, these proteins stimulate a 24-30-fold expansion of HSCs following 10 days of culture of highly enriched mouse stem cells. Ongoing work indicates that a similar “cocktail” of five growth factors in serum- free medium will support a ~30- fold expansion of human hematopoietic cord blood stem cells, and we are currently collaborating with several clinical laboratories to carry out preclinical and eventually clinical studies on cord blood HSC expansion.
Supportive stromal cells for hematopoietic stem cells Hematopoietic stem cell environments or niches are very important in determination of HSC self-renewal and differentiation. Fibroblasts, endothelial cells, and osteoblasts have been postulated as important constituents and regulators of HSC niches in the bone marrow. We are interested in characterizing additional cell types that contribute to regulation of HSC microenvironments and as noted we uncovered one such population from fetal liver. Megan Kaba and Alec Babic are studying different subpopulations of fetal liver immature T lymphocytes and their potential to interact with and regulate stem cell self renewal and expansion during fetal and adult hematopoiesis. Specifically, Megan and Alec are performing coculture experiments to further characterize which populations of fetal liver cells are able to enhance the ex vivo expansion of long-term repopulating adult bone marrow HSCs. Of interest are the embryonic day 15 fetal liver CD3+ Ter119- cells that express the T cell beta or gamma-delta receptors. Characterization of a specific type of supportive cell will aid in the identification of yet other new growth factors that stimulate HSC expansion.
Song Chou, together with Yachao Liu, are also trying to identify the
stromal cells that support HSC expansion in fetal liver. By using real-time
PCR, Song discovered that fetal liver CD3+ cells not only are highly enriched
for Angptl3 and IGF2 mRNAs, but also for stem cell factor (SCF), The membrane-
anchored ligand for the c-kit tyrosine kinase receptor, and TPO mRNAs.
Song also sorted E15.5 fetal liver cells according to surface expression
of SCF and found SCF+ cells are enriched for the mRNAs encoding Angptl3,
TPO and IGF2, in comparison to SCF- cells. Since these four hormones form
a complete set of growth factors that are able to significantly expand
HSCs ex vivo, it is likely that subpopulation(s) of fetal liver CD3+ cells
are able to secrete all four crucial growth factors and are able to support
HSC expansion in fetal liver in the absence of other added cytokines.
Plasma membrane SCF binds to its receptor, c-Kit, in adjacent cells. Since
all HSCs in fetal liver express c-Kit, these stromal cells may be located
in close proximity to HSCs and interact with HSCs through SCF. Song and
Yachao are developing methods to further purify these potential HSCs stromal
cells and to identify additional signal molecules emanating from these
cells for HSC expansion.
Role of the prion protein in self- renewal of hematopoietic stem
cells Among the other proteins specifically expressed by Day
15 fetal liver CD3+ cells was the prion protein (PrP), a glycosylphosphatidylinositol
(GPI)- anchored cell surface protein; despite many years of research the
normal function of PrP was unknown. Chengcheng Zhang surmised that PrP
would also be expressed on long-term repopulating hematopoietic stem cells
and initiated a collaboration with Professor Susan Lindquist and her PhD
student Andrew Steele. Not only did they quickly confirm this hypothesis,
they went on to show that HSCs from PrP -/- bone marrow exhibit impaired
activities in serial transplantation experiments. Most strikingly, ectopic
expression of PrP in PrP -/- bone marrow cells rescued the defects in
hematopoietic engraftment. Therefore, PrP is a novel marker for HSCs and
supports their self-renewal during successive bone marrow transplantations.
More recently Megan Kaba, working with Andrew, showed that aging PrP -/-
mice have fewer HSCs than normal, and also exhibit other hematopoietic
defects. Thus PrP supports HSC self-renewal during normal adult life.
Song Chou is trying to determine the molecular function of PrP. PrP might
be the coreceptor for a hormone affecting HSC activity, possibly concentrating
this as yet unidentified molecule on the cell surface and/or presenting
it to the signaling receptor(s). Alternatively, PrP might interact with
proteins in the BM extracellular matrix or on the surface of stromal cells,
and possibly support retention of HSCs within the bone marrow microenvironment.
MicroRNAs (miRNAs) are ~22-nt non-coding RNAs that can play important roles in development by targeting the messages of protein-coding genes for cleavage or repression of productive translation. As shown by the Bartel laboratory and others, humans have over 300 genes that encode miRNAs, an abundance corresponding to almost one percent of protein-coding genes. Based on the evolutionary conservation of many miRNAs among different animal lineages, it is reasonable to suspect that some mammalian miRNAs might also have important functions during development.
MicroRNAs that modulate hematopoiesis As a first step
towards testing the idea that miRNAs might play roles in mammalian development,
and more specifically hematopoiesis, Chang- Zheng Chen in collaboration
with David Bartel, cloned about 100 unique miRNAs from mouse bone marrow.
Three, miR-181, miR-223, and miR-142s, were exclusively or preferentially
expressed in hematopoietic tissues. miR-181 was very strongly expressed
in thymus, the primary lymphoid organ, which mainly contains T-lymphocytes.
Mature miR-181 expression in the bone marrow cells was up-regulated in
differentiated B-lymphocytes. Using retrovirus vectors he developed, Chang-
Zheng ectopically expressed miR-181 in a population of bone marrow hematopoietic
stem and progenitor cells. This led to an increased fraction of B-lineage
cells both in tissue-culture differentiation assays and in transplanted
adult mice; there was a corresponding decrease in CD-8+ T cells. These
and other results with other miRNAs indicate that microRNAs are components
of the molecular circuitry controlling mouse hematopoiesis and suggest
that other microRNAs have similar regulatory roles during other facets
of vertebrate development. Some of this work is being done in Chang- Zheng’s
own laboratory at the Stanford University School of Medicine.
Beiyan Zhou has used microRNA microarrays developed in the Bartel laboratory and Northern Blot analysis to identify several miRNAs specifically upregulated in isolated populations of thymic and splenic hematopoietic cells: B cells, immature CD-4- CD-8- and CD-4+ CD-8+ T cells, as well as in more mature thymic CD-4- CD-8+ and CD-4+ CD-8- T cells. She has confirmed these results by Northern blotting. Based on these results, five microRNAs (miR-195, miR-150, miR-106, miR-181a, and miR142) were chosen for further analysis. miR-150 and miR-181a were highly expressed in the thymus and spleen, the major secondary lymph organ, leading to our hypothesis that these two miRNAs are involved in lymphopoietic regulation.
To study the potential regulatory roles of these miRNAs in T cell function Beiyan and Stephanie Wang, a UROP student, have used retroviral infection of two clonal CD8+ T-cell lines to establish 8 stable clonal CD8+ T-cell lines that ectopically express either miR-106, miR-142, miR-150, or miR-195. She is currently establishing CD4 clonal cell lines that overexpress these selected miRNAs. These infected T-cell clones will be used for experiments to examine the effects of miRNAs on mature T-cell function. Meanwhile, Beiyan also discovered a group of microRNAs that have been up or down regulated during T cell activation. The regulatory functions of these microRNAs are under investigation now.
Beiyan also studied the expression patterns of these selected microRNAs in detail during several various developmental stages of both B and T cells. miR-150 is mainly expressed in the lymph nodes and spleen, and is highly upregulated during the development of mature T and B cells. In particular, expression of miR-150 is sharply upregulated at the immature B cell stage. In contrast, expression of miR-181 is sharply downregulated during B cell development, and several years ago Chang- Zhen Chen showed that overexpression of miR-181 in hematopoietic stem/ progenitor cells led to a significant increase in production of mature B cells. To explore the roles of miR-195 and miR-150 in lymphopoiesis Beiyan generated retroviral constructs that express each of these miRNAs. Ectopically expressed miR-150 (but not miR 195) elicited a significant inhibition of B cell development, and had little effect on other hematopoietic lineages. Further analysis showed that miR 150 overexpression did not affect B cell lineage commitment, as evidenced by unaltered pro-B cell numbers both 4 and 16 weeks after transplantation. Overexpression did block the pro-B to pre-B transition in the bone marrow. We hypothesize that premature expression of miR-150 in hematopoietic stem/progenitor cells inhibits production of proteins specifically required for early B cell development.
Several potential miR-150 targets were predicted by TargetScan screening. The mechanism of miR-150 regulation in B cell development is being further investigated by studying these potential targets, in particular B lineage specific transcription factors. One of the important transcription factor targeted by miR-150 is Myb, which is also an important regulator for B cell development. Mice with the myb gene selectively deleted in the B cell lineage have a B cell development deficiency. In order to test the direct regulatory relation between miR-150 and Myb, Beiyan is generating a knock in mouse model with potential miR-150 target sites mutated in the Myb 3’UTR region. This will provide us with important information on the role of miR-150 regulation of myb gene expression during B cell development. Other potential targets of this microRNA at this stage are also being investigated. Furthermore, Andrew Shie, a UROP student is currently working with Beiyan to further investigate the regulatory role of miR-150 in early B cell development. Beiyan is also trying to understand the potential role of a group of microRNAs that are up or down regulated during the final B cell development, activation.
MicroRNAs affecting drug- resistance of leukemias Acute lymphoblastic
leukemia (ALL) is one of the most common malignancies of children and
young adults. Ai Kotani aims to identify and characterize specific miRNAs
that are critical in the development and progression of MLL related leukemias,
an ALL that shows poor prognosis. In MLL related ALL, expression of many
miRNAs is downregulated, raising the possibility that downregulation of
some miRNAs plays a critical role in the pathogenesis of this disease.
miR-128 is one of these. Ai Kotani and Daon Ha, a UROP student, studied
RS4; 11 cells, a cell line derived from a MLL-AF4 ALL patient in which
a balanced translocation between the MLL and AF4 genes has occurred. RS4;
11 cells have a novel A to G point mutation in the miR-128b gene segment;
this mutation is transcribed in the primary miR- 128 transcript and blocks
the processing of the miR-128 precursor to mature miRNA. Like MLL related
ALL, the RS4; 11 cell line is resistant to glucocorticoid- induced apoptosis.
Ai and Daon showed that overexpression of the wild- type miR-128 gene
in these cells restored glucocorticoid- induced apoptosis. Target genes
directly downregulated by miR-128b include MLL, AF4, and both MLL-AF4
and AF4-MLL fusion genes. miR-221 normally downregulates the CDKN1B gene,
which encodes the p27 cell cycle inhibitory protein. Overexpression of
miR-221 also restored glucocorticoid- induced apoptosis to RS4; 11 cells
and functioned synergistically with miR-128. These results demonstrate
that downregulation of miR-128b and miR-221 induces glucocorticoid resistance,
and that restoration of their levels is a promising therapeutic in MLL-AF4
ALL.
Currently Ai and Doan are trying to understand the mechanisms by which
these miRNAs affects steroid sensitivity. They are also studying another
miRNA that is downregulated in MLL-AF4 ALL and that induces specific phenotypic
changes in leukemic cells.
MicroRNAs that modulate adipogenesis Following the work of I-hung Shih, a former postdoctoral fellow in the Bartel lab who worked closely with us, Huangming Xie is examining the role of miRNAs in adipogenesis using several adipocyte cell culture differentiation systems. He profiled miRNA expression during in vitro adipogenesis of the preadipocyte 3T3-L1 cells using miRNA microarrays and validated by RT-PCR eight miRNAs that are significantly upregulated and four that are downregulated. Similar changes in miRNA expression were observed by comparison of mature primary adipocytes and enriched primary preadipocytes. He also profiled miRNA expression in purified mature adipocytes and compared miRNA profiles in epididymal adipocytes from normal and leptin deficient or diet-induced obese mice. These changes are likely associated with the chronic inflammatory environment in obese adipose tissue as they were mimicked by TNFalpha treatment of differentiated adipocytes. Huangming and Lei Sun, a new postdoctoral fellow, are investigating the role of these candidate miRNAs in adipogenesis and adipocyte functions in vitro and in vivo.
Ectopic expression of two adipocyte-enriched miRNAs in preadipocytes accelerated adipogenesis, as measured both by the upregulation of many adipocyte-important genes including adiponectin and the key transcription factor PPARgamma, and by an increase in triglyceride accumulation at an early stage of adipogenesis. Currently, Huangming is investigating the mRNA targets of these candidate miRNAs and examining the regulation of miRNA expression using bioinformatics tools followed by experimental validation. Huangming and Lei has also compared miRNA profiles between white and brown adipose tissue and these differences are likely attributed to their different origins, niches and contrasting lipid metabolic functions.
Together, these studies present a comprehensive view of miRNA dynamics in adipocyte development and function and provide an important first step towards construction of the entire RNA regulatory network underlying fat cell development and adipose dysfunction in obesity. These findings have important implications for the understanding the link between chronic inflammation and obesity with insulin resistance. An understanding of the role of miRNAs in adipose biology, obesity and related medical complications such as insulin resistance may lead to novel RNA-based therapies that complement current anti-obesity treatments.
A cluster of five miRNAs, the mir-17-92 cluster, is frequently upregulated
in cancers, but its normal mRNA targets are unknown. A new postdoctoral
fellow, Bill Wong, is studying the context-dependent role of these miRNAs
in regulating both hematopoietic and adipocyte differentiation. Several
of these studies involve generation and analysis of mice with lineage-
specific ablations of this gene cluster.
Muscle differentiation and function Prakash Rao started his studies in the laboratory by identifying three microRNAs, miR-1, miR-133 and miR-206 that are upregulated during the differentiation of the C2C12 myoblast cell line into myotubes. In a collaborative study, Prakash (along with Dr. Roshan Kumar of the Young lab at Whitehead) identified the myogenic transcription factors responsible for their specific regulation during myogenesis.
The observation that miR-1, miR-133 and miR-206 are induced during the differentiation of a myoblastic line have led them to consider their potential mRNA targets in muscle differentiation. In particular, they have focused on the ability of miR’s to target components of conserved regulatory cascades that inhibit myogenesis, and on the regions within the 3` UTRs of these mRNAs that possesses miR-1 binding sites. Prakash and Greg Hyde, a visiting scientist, showed that miR-1 can inhibit reporter genes bearing regions of some of these potential target mRNAs. Overall, these studies point to the importance of miR-1 in promoting the myogenic program. Consistent with these hypotheses, Prakash, with able assistance from another UROP, Lauren Shields, has demonstrated the ability of miR-1 to promote differentiation of a muscle-derived tumor cell line. Taken together with the observation that miR-1 levels are lower in muscle-derived tumors (data from Janet Shipley at the Institute for Cancer Research in the UK), it is likely that miR-1 is a tumor suppressor in rhabdomyosarcoma cells. Prakash plans to test these ideas using in vivo systems.
Prakash, in collaboration with Drs Robert Blelloch (UCSF) and Rudolf
Jaenisch, is also studying the impact of the global loss of all microRNAs
by deleting (specifically in muscle tissue) a gene required for microRNA
biogenesis. Lack of dgcr8 leads to a reduction in mature microRNAs in
the heart and skeletal muscle and leads to the development of heart failure.
Various studies are underway in these mice to gain greater insight into
the reason underlying their mortality. These studies complement his efforts
at cataloging all microRNAs expressed in muscle tissue. This work was
done using deep sequencing of small RNAs and is being carried out jointly
with Rosaria Chiang in the Bartel lab.
In 1995 we cloned adiponectin, originally called Acrp30, as a novel adipocyte- specific secreted protein hormone. Adiponectin addition potently elevates fat and glucose catabolism by muscle, enhances glycogen accumulation in muscle, and inhibits gluconeogenesis in liver. Mutations in the adiponectin gene are linked to development of adult- onset diabetes and the levels of adiponectin in serum are reduced in obese and diabetic patients and mice. Circulating adiponectin levels negatively correlate with human plasma triglyceride and fasting insulin levels and several clinical studies showed persons with low adiponectin levels are more likely to develop type II diabetes mellitus and cardiovascular disease. This data suggests that adiponectin is a potential genetic determinant of insulin sensitivity.
Adiponectin has four domains: a cleaved amino-terminal signal sequence, a region without homology to known proteins, a collagen-like region, and a globular segment at the carboxy-terminus. The three-dimensional structure of the globular domain is strikingly similar to that of TNF -alpha even though there is no homology at the primary sequence level. Like TNF -alpha the globular domain forms homotrimers, and intermolecular disulfide bonds generate hexameric and high molecular weight Adiponectin species.
In collaboration with the Ruderman laboratory at B. U. Medical School,
we showed several years ago that treatment of rat striated muscle with
trimeric adiponectin led to phosphorylation and activation of AMP-activated
protein kinase (AMPK), an enzyme that when activated causes increases
in muscle fatty acid oxidation, glucose uptake and oxidation, and insulin
sensitivity. Adiponectin- mediated activation of AMPK caused phosphorylation
and thus diminished activity of acetyl CoA carboxylase, a corresponding
decrease in the concentration of malonyl CoA, and a corresponding increase
in long- chain fatty acid oxidation. In addition, adiponectin caused an
increase in glucose uptake by muscle. Both in vivo and in muscle culture
adiponectin most likely exerts its actions on muscle fatty acid oxidation
by inactivating ACC, via activation of AMPK and perhaps other signal transduction
proteins.
T- cadherin and other receptors for adiponectin Christopher Hug, assisted by Jin Wang, used an expression cloning strategy to identify T- cadherin as a receptor for hexameric and high molecular weight forms of adiponectin. T-cadherin is highly and specifically expressed in the vasculature, where it is predominantly found in endothelial and smooth muscle cells in the blood vessel intima. At its C-terminus T-cadherin is attached to the plasma membrane via a GPI anchor. Chris’ studies indicate that T-cadherin is the major adiponectin binding protein in the body, as deletion of T-cadherin results in a many-fold increase in the level of all isoforms of adiponectin in the circulation. Immunohistochemical localization of adiponectin demonstrated that mice lacking T-cadherin had no detectable binding of adiponectin to the vascular endothelium, in contrast to wild-type animals that had substantial binding of adiponectin to the endothelium. T- cadherin is upregulated following vascular injury and Chris hypothesizes that, by binding to adiponectin, it may play a role in atherosclerosis progression as well as blood vessel formation and endothelial cell function.
Furthermore, T-cadherin null mice demonstrate hepatic insulin resistance,
a phenotype virtually identical to that of adiponectin knockout animals.
In conjunction with Dr. Gerry Shulman’s lab at Yale Medical School,
Chris has characterized the metabolic and physiologic abnormalities of
mice lacking T-cadherin, which may mimic those of the metabolic syndrome.
Using the hyperinsulinemic euglycemic clamp technique on mice fed a high-fat
diet for three weeks, they demonstrated that after an overnight fast there
was no difference in insulin stimulated whole body glucose uptake, glycolysis,
and glycogen synthesis. However, hepatic glucose production rates in T-cadherin
deficient animals during the clamp were significantly higher than those
of the control animals. Thus, T-cadherin deficient mice demonstrate hepatic,
but not peripheral, insulin resistance. These results confirm that T-cadherin
is a bonafide receptor for high-molecular weight forms of adiponectin,
and that loss of T-cadherin causes a metabolic phenotype similar to that
reported for loss of adiponectin.
Currently Chris is determining the role of T-cadherin in adiponectin activation of the AMPK and NF-kB signal transduction pathways, and is studying the downstream pathways activated by adiponectin binding to T-cadherin. As T-cadherin lacks a transmembrane domain, and is thus not likely to be a signaling receptor, Chris and Jin are also cloning other cell surface adiponectin receptors including those that directly activate AMPK. Finally, as adiponectin shares significant structural but not sequence similarities with TNF-alpha, Chris is testing all the known TNF-alpha superfamily receptors for possible binding to recombinant adiponectin. These studies will continue in Chris’ own laboratory at Children’s Hospital Harvard Medical School.
Activation of AMP kinase by adiponectin Adiponectin has important roles in enhancement of insulin sensitivity and these beneficial effects are closely associated with the activation of AMP-activated protein kinase (AMPK) in muscle and liver. How adiponectin activates AMPK is not known. Interestingly, AMPK activation by adiponectin is accompanied by an increase in concentration of 5’AMP, which implies the presence of signal transduction proteins in a pathway connecting the adiponectin signal with an increase in 5’AMP level. Qingqing Liu is determining, first, what metabolic or signaling pathway(s) downstream of adiponectin receptors leads to this rise in 5’AMP. Second, she is determining whether this rise in 5’AMP is necessary for activation of AMPK. Activation of long-chain fatty acids to the CoA derivative is one important metabolic process that directly generates 5’AMP, and at least some Acyl CoA synthase isoforms are on the plasma membrane. She therefore proposes that one or more of Acyl CoA synthase enzymes are directly coupled to adiponectin receptors, possibly to T- cadherin. Her preliminary data show that free fatty acids, essential substrates for the production of 5’AMP by acyl-CoA synthases, are required for AMPK activation by adiponectin in C2C12 myocytes. She further demonstrated that Acsl1 and FATP1 are the two major isoforms expressed in skeletal muscle. She is currently determining which acyl-CoA synthases are involved in adiponectin activation of AMPK, using siRNAs that specifically target individual Acsl or FATP family proteins. Moreover, she is using immunofluorescence and other approaches to test the hypothesis that adiponectin enhances acyl CoA synthase activity by stimulating their translocation from intracellular compartments to the plasma membrane.
Although the effects of adiponectin on glucose and lipid metabolism in
liver and skeletal muscle were reported to be mediated by two receptor
isoforms, AdipoR1 and AdipoR2, work in several laboratories including
our own failed to confirm these as receptors. As T-cadherin is attached
to the plasma membrane by a glycosylphosphatidylinositol anchor and lacks
any transmembrane or cytoplasmic domain, it is also unlikely to be a signaling
receptor. Qingqing intends to identify adiponectin receptors using a biotin-label-transfer
method. These studies are essential to understand the mechanisms underlying
adiponectin activity and potentially exploit these to develop new strategies
to treat metabolic diseases.
Activation of AMP kinase by IL-6 AMPK is composed of three subunits - the alpha kinase subunit that undergoes regulated phosphorylation, the gamma subunit that binds AMP, and the beta subunit that is thought to act as a scaffold that binds to both the alpha and gamma subunits. Cellular and physiological stresses that deplete ATP such as nutrient deprivation, hypoxia, ischemia, and exercise in muscle all lead to activation of AMPK. Additionally adiponectin and cytokines including IL-6, CNTF, and leptin have also been shown to activate AMPK. It is not exactly clear how this activation occurs. Kelly Wong is dissecting the mechanism by which IL-6 activates AMPK and will determine if other cytokines use a similar mechanism. Through this project, Kelly also hopes to find a connection to the elusive adiponectin signaling receptors. Kelly is also working on determining the physiological role of the beta-subunit and its importance in glycogen binding. To this end he has undertaken an in vivo loss of function approach to create and analyze mice deficient in the beta-subunit, and also mice that harbor a specific germ line point mutation in the beta-gene that deletes its glycogen binding ability.
Adiponectin and development of mammary and other cancers
Both in mice and humans the serum levels of adiponectin are inversely
correlated with adipose mass and body mass index (BMI). Epidemiological
studies have demonstrated that serum adiponectin levels have an inverse
association with breast cancer risk. One study indicated that low serum
adiponectin concentrations, such as occur during obesity, were associated
with large tumors and tumors of high histological grade. Moreover, adiponectin
inhibits growth of several cell types, including vascular smooth muscle
cells and several breast cancer lines although the signal transduction
pathways it activates and the mechanism(s) by which it exerts these anti-proliferative
effects are unknown. Therefore, it is important to reveal the role of
adiponectin in breast cancer formation and progression. Yutong Sun hypothesizes
that low serum level of adiponectin will accelerate mammary tumor formation
and lead to larger tumors in mice, and that adiponectin can inhibit breast
tumor cell proliferation, survival and migration/invasion.
In order to determine whether adiponectin deficiency can decrease the
latency and/or increase the incidence of breast cancer formation in mice,
Yutong is breeding breast cancer mouse models (MMTV-Her2/neu transgenic
mice, and MMTV-Cre/Flox neo NeuNT mice from Dr. William Muller, McGill
University) into an adiponectin -/- background. He will also expose adiponectin
-/- and wild- type mice to carcinogens, such as NMU or DMBA. Furthermore,
employing mammary tumor cell implantation, he will examine whether adiponectin
deficiency affects tumor growth, angiogenesis and metastasis in mice.
Currently, he is exploring whether purified recombinant adiponectin has
any effect on tumor cell proliferation, survival and migration/invasion
Adiponectin paralogs Guang William Wong, with the assistance
of Claire Kitidis, used multiple genomic approaches to identify a family
of ten highly conserved human and mouse adiponectin paralogs. These are
designated as C1q/TNF-alpha related proteins (CTRP)-1 to 10.
Of all the CTRP paralogs, the highly- conserved CTRP9 shows the highest degree of amino acid identity to adiponectin in its globular C1q domain. CTRP9 is expressed predominantly in adipose tissue and female expresses higher levels of the transcript compared to male mice. CTRP9 is a secreted glycoprotein with multiple posttranslational modifications in its collagen domain that include hydroxylated prolines and hydroxylated and glycosylated lysines. It forms and is secreted as multimers (predominantly trimers) from transfected cells and circulates in the mouse serum. Furthermore, CTRP9 and adiponectin are secreted as hetero-oligomers when co-transfected into mammalian cells, and in vivo, adiponectin/CTRP9 complexes can be reciprocally co-immunoprecipitated from the serum of adiponectin and CTRP9 transgenic mice. Biochemical analysis demonstrates that adiponectin and CTRP9 associate via their globular C1q domain, and this interaction does not require their conserved N-terminal cysteines or their collagen domains. Furthermore, using gel filtration chromatography combined with co-immunoprecipitation analysis, they showed that adiponectin and CTRP9 form heterotrimers. Because different oligomeric forms of adiponectin have distinct biological activities, identification of CTRP9 that can heterotrimerizes with adiponectin impacts the study of adiponectin function. In cultured differentiated myotubes CTRP9 specifically activates AMPK, Akt, and p44/42 MAPK signaling pathway. And adenovirus-mediated overexpression of CTRP9 significantly lowered serum glucose levels in obese (ob/ob) mice compared to controls. Collectively, these results suggest that CTRP9 is a novel adipokine and further study of CTRP9 will yield novel mechanistic insights into its physiologic and metabolic function.
Guang and Claire showed that CTRP1, CTRP2, CTRP3, CTRP5, and CTRP7 transcripts are expressed predominantly by adipose tissue, with gender-biased expression for CTRP3 and CTRP6. Expression of CTRP1, CTRP2, CTRP3, CTRP6, and CTRP7 varied according to the age of obese (ob/ob) mice, with significant upregulations in younger mice. Rosiglitazone, an agonist of the master transcriptional regulator PPARgamma, increased the expression of CTRP1 while decreased CTRP6 transcripts in vivo.
All CTRPs are secreted glycoprotein, with CTRP1, CTRP2, CTRP3, CTRP5, and CTRP6 found circulating in the serum; their levels vary according to the gender and genetic backgrounds of mice. Intriguingly, serum levels of CTRP1 and CTRP6 are increased in adiponectin-null mice. Like adiponectin, CTRP1, CTRP2, CTRP3, CTRP5, CTRP6, and CTRP10 proteins form trimers as their basic structural unit. CTRP3, CTRP5, CTRP6, and CTRP10 are further assembled into higher order oligomeric forms via N-terminal disulfide bonding. In addition to forming homo-oligomers, systematic co-transfections followed by co-immunoprecipitations revealed that CTRP1/CTRP6, CTRP2/CTRP7, and adiponectin/CTRP2 can also be secreted as hetero-oligomers. In sum, these molecular and biochemical data provide an important and valuable framework for the functional analysis of this family of proteins using in vitro and in vivo approaches.
An understanding of the natural metabolic functions of these hormones
will likely emerge from analysis of the CTRP- overexpressing transgenic
mice and CTRP gene knock- out mice Guang is now generating. Guang is also
using expression cloning strategies to identify the receptors for these
novel proteins. This discovery of a family of adiponectin paralogs has
implications for understanding the control of energy homeostasis and could
provide new targets for pharmacologic intervention in metabolic diseases
such as diabetes and obesity. Much of this work is continuing in Guang’s
new laboratory at the Johns Hopkins School of Medicine.
Protease cleavage and release of the extracellular domain (ECD, "ectodomain shedding") of a multitude of transmembrane proteins has been linked to the activation of many signaling pathways including the MAPK pathway. Cleavage of the ECD is mostly carried out by metalloproteases (MMPs) of the ADAM family (“a disintegrin and metalloprotease”). ECD cleavage is often followed by and is a prerequisite for intramembranous cleavage of the intracellular domain (ICD) of the same protein by gamma-secretase; some of the cleaved ICDs translocate to the nucleus, where they may regulate gene transcription. Membrane-spanning pro-hormone ligands of the epidermal growth factor receptor (HER) family are well-studied examples of proteins that undergo ectodomain shedding and are physiologically important in many cellular contexts in organisms from Drosophila to mammals. But how the ectodomain cleavage machinery is regulated is largely unknown, as only a few specific stimuli that induce ectodomain shedding have been identified. Prolonged activation of the cardiac beta-adrenergic receptor leads to HB-EGF-cleavage, release of soluble HB-EGF, and development of cardiac hypertrophy. Andreas Herrlich showed that another HER-ligand, neuregulin1beta (NRG1beta), is cleaved by an MMP in response to hypertonic stress and subsequently activates EGF- family receptors in an autocrine fashion. This signaling step leads to MAPK activation followed by enhanced expression of genes encoding water channels (aquaporins). Regulation of ectodomain cleavage could occur at least two levels - at the level of the MMP or via covalent modifications of the target protein, such as phosphorylation or ubiquitination on the cytosolic face.
Andreas, with the assistance of Eva Klinman, an MIT undergraduate, is
cloning novel genes that regulate ectodomain shedding using a high-throughput
expression cloning strategy. They can detect cleavage of all chosen HER-ligands
either by hypertonic stress, phorbol ester addition, or stimulation with
lysophosphatidic acid in a FACS-based assay using mouse or human cell
lines stably expressing one of the chosen pro-hormone ligands. The ligands
are tagged at the extracellular domain with one of several epitope tags;
at their cytosol-facing C- termini the proteins have been fused with EGFP.
The extracellular epitope of the transmembrane pro-hormone ligand is detected
with a fluorochrome-coupled (red) anti-epitope antibody, while the intracellular
domain of the EGFP- fusion is detected by green fluorescence. Stimulation
of cleavage results in a decrease of the red to green fluorescence ratio,
while inhibition of basal or induced cleavage is reflected by an increase
in this ratio. Andreas’ initial studies with this system showed
that, when expressed in mouse lung epithelial cells, ectodomain cleavage
of these three EGF ligands is specifically triggered by different stimuli
and involves different PKC isoenzymes. Studies utilizing inhibitors of
protein kinase C isoenzymes or metalloproteinase inhibition by batimastat
showed that different regulatory signals are used by different stimuli
and EGF substrates, suggesting differential effects that act on the substrate,
the metalloproteinase, or both. Andreas and Eva are now using this assay
system for a 96 well plate high-throughput shRNA gene knockdown screen.
Currently they are in the early stages of testing the effect of shRNA
constructs targeting about 5% of all known mammalian kinases and phosphatases
on TPA-induced ectodomain cleavage of TGF-alpha. By stimulating cleavage
to only about 50% of total possible cleavage they can detect both, decreases
(inhibitors of cleavage) and increases (activators of cleavage) in red:
green fluorescent ratio in the same screen. Once candidate genes are identified
they will also be tested in the context of the other physiological cleavage
stimuli, hypertonic stress and GPCR stimulation.