Richard A. Anderson, Ph.D. Director Molecular and Cellular Pharmacology Program Professor Department of Pharmacology raanders@wisc.edu

Research Interests: The Role of Signal Transduction in Control of Cellular Processes. 

PIP2 is directly synthesized by particular members of the family of phosphatidylinositol-phosphate kinases (PIPKs) and PIP2 occupies an essential role in PI signaling by directly regulating cellular functions that include cell proliferation, secretion and vesicular trafficking, cellular morphogenesis, and cell motility. In addition, PIP2 is also a key signal transduction molecule as the precursor for many other second messengers, such as PI3,4,5P3 and soluble inositol polyphosphates.

All of the different PIP kinase (PIPK) isoforms target to discrete subcellular compartments, such as nuclear speckles, focal adhesions and intercellular membrane compartments as the example shown in Figure 1 for the PIPKI isoforms. We have discovered that different PIPK family members are targeted to compartments through unique protein-protein interactions. We hypothesize that these interactions between the PIPKs and targeting proteins result in the spatial and temporal generation of PIP2 that in turn regulates specific cellular functions (Fig. 2). Most of the proteins that directly interact with the PIP kinases and spatially target the kinase with in cells are themselves PIP2 effectors. The interaction between the PIP kinase that generates PIP2 and its effector is a mechanism to bestow PIP2’s specificity as a lipid messenger by spatially generating PIP2 at cellular compartments rich in its effectors. In turn, if the interaction between the PIPKs and PIP2 effectors is the basis for signaling and functional specificity, then the interactions between the PIPK and PIP2 effectors should be tightly regulated in order to control the signaling pathway in which a specific PIP2 effector functions.  We now have multiple examples of specific interactions between PIPKs and PIPn effectors. These interactions are governed by signals that regulate the cellular pathway controlled by that PIPn effector.

These observations have allowed us to define the role of PIPKs in specific cellular functions by defining their interacting partners. The discovery of PIPK interacting proteins that are also PIP2 effectors has lead to a conceptual understanding of how PIP kinases modulate a wide variety of cellular functions. The laboratory focuses on two broad areas of investigation. 

The Role of PIP Kinases in Epithelial Cancer Metastasis. The majority of cancers (~75%) are derived from epithelial cells. For cancers to metastasize, cells must first detach from the original tumor and migrate into the surrounding tissue. In cancers of epithelial origin, loss of E-cadherin based cell-cell contacts results in a loss of cell polarity and transforms cells into a migratory phenotype; this is a hallmark of cancer progression, and a key step in detachment of cells from the initial tumor (see Fig. 3). Upon transformation to the mesenchymal or migratory phenotype, chemoattractants such as epidermal growth factor stimulate cell invasion and migration into the vasculature or lymph system. We have shown that phosphoinositide signaling plays vital roles in both the assembly/disassembly of E-cadherin junctions and growth factor stimulated directional migration and invasion of mesenchymal cells. Both of these processes are controlled by PIPKIg isoforms and these functions position this enzyme as a key regulator of the metastatic process.

Our work has demonstrated that PIPKIg isoforms modulate epithelial cell polarity through control of E-cadherin assembly into adherens junctions as well as E-cadherin trafficking to and from the plasma membrane. PIPKIg isoforms regulate E-cadherin function via a direct interaction with E-cadherin, and then in an isoforms specific manner, recruit specific interacting proteins to the E-cadherin-PIPKIg complex. These additional interactors including the clathrin adaptor complexes (AP complexes), the sorting nexins and others, are PI4,5P2 effectors that drive the trafficking and assembly of E-cadherin.

Strikingly, when epithelial cells transition into a mesenchymal or migratory phenotype, the same PIPKIg isoform targets to focal adhesions through a direct interaction with talin, which in turn regulates the dynamics of focal adhesion assembly. Most significant, the focal adhesion targeted PIPKIg isoform is specifically regulates growth factor-stimulated chemotaxis and invasion.  These discoveries support a pivotal role for PIPKIg in the progression and metastasis of epithelial cancers. To validate this work we are establishing mouse models of breast cancer progression where progression and metastasis are driven by growth factor receptors.  In these models the PIPKIg isoforms will be knocked out or mutants knocked in and the impact on the metastasis of the tumor will be quantified.
To translate this work, we have begun to explore putative roles of PIPKIg in breast cancer progression. This research emphasis takes two approaches: first, using a breast cancer patient tumor tissue microarray and database we are correlating changes in PIPKIg isoform expression with patient outcomes and other biomarkers known to have roles in progression or suppression of metastasis. The second approach focuses on mouse models to determine the role played by PIPKIg in metastasis in vivo. In collaboration with David Huntsman at the Vancouver Cancer Center, using a large and well-characterized breast tumor tissue microarray, we have discovered that the expression level and/or membrane targeting of PIPKIg correlates well with tumor invasion and poor patient outcome in breast cancers using a large patient tissue microarray. This suggests that PIPKIg splice variants may be unique makers for the metastatic potential of breast cancers and other cancers of epithelial origin. This broad focus is funded by three NIH grants: R01 CA104708, R01 GM057549, and P30 CA014520-AV-130.

Nuclear phosphoinositide signaling pathways. The cell nucleus is a highly organized structure and this organization is critical for the normal functioning of the nucleus. The role of the PIP kinases in the regulation of nuclear signaling represents a fundamental and largely unexplored area of biology. We have discovered that PIPK isoforms spatially target to structures called nuclear speckles and PIP2 is generated at these same sites (Fig. 4). Nuclear speckles at the EM level are called interchomatin granule clusters and these structures contain proteins that play roles in transcription and processing of pre-mRNA. Our hypothesis is that the PIPKs localize to speckles where they generate PIP2 and derived messengers that regulate nuclear events in and around speckles. These possibilities are exciting, but also puzzling; since phosphoinositides and PIPKs are found in nuclei at sites apparently devoid of membrane structures. 

Two PIP kinases are targeted to nuclear speckles, PIPKIa and PIPKIIb. We have shown that PIPKIa and PIPKIIb function in distinct signaling pathways. Both of these enzymes generate PI4,5P2 from PIP substrates but PIPKIa uses PI4P as a substrate whereas PIPKIIb uses PI5P as the substrate. PIPKIa generates PI4,5P2 as the lipid mediator. Interestingly, in the PIPKIIb pathway, PI5P is thought to be the key lipid messenger and PIPKIIb regulates this messenger by conversion to PI4,5P2. Since PIPKs targeting protein partners are often PIP2 effectors, this suggested that specific protein-protein interactions are required for PIPK nuclear speckle targeting. Identification of these targeting and effector proteins would lead to defining the function of the PIPKs within the nucleus. Indeed, the identification of PIPKIa and PIPKIIb interacting nuclear proteins has lead to the discovery of two novel signaling pathways in nuclei. 

PIPKIa regulates 3’-end processing of pre-mRNAs. The PIPKIa directly interacts with a novel poly(A) polymerase, and we called this enzyme Star-PAP for nuclear Speckle Targeted PIPKIa Regulated-Poly(A) Polymerase. PIPKIa and Star-PAP co-localize at nuclear speckles and Star-PAP polymerase activity. Remarkably Star-PAP activity is specifically stimulated by PI4,5P2, the product of PIPKIa. Poly(A)polymerases (PAPs) have multiple functions within cells including the regulation of RNA quality control and degradation by the exosome, polyadenylation of mRNAs in the cytosol, and polyadenylation of pre-mRNAs by the RNA polymerase II (RNA Pol II) complex.

Several decades of literature have instituted a dogma that RNA Pol II assembles a canonical PAPa that generically polyadenylates all pre-mRNAs. We have shown that Star-PAP functions in the RNA Pol II complex to process select pre-mRNAs. Our data revealed that Star-PAP integrates into a RNA Pol II complex in response to specific stimuli and in the absence of detectable canonical PAP. This supports a model for the assembly of the polyadenylation machinery where either Star-PAP or canonical PAP, but not both, can integrate into the RNA Pol II transcriptional complex.

Analysis of gene expression by microarray and quantitative real-time RT-PCR demonstrated that Star-PAP is required for the expression of a relatively small fraction of mRNAs. Stress response pathways activate the Star-PAP dependent genes that we have characterized, and these genes have been reported to play roles in Alzheimer’s disease, cardiovascular disease, the immune response, and aspects of cancer progression. Of these mRNAs, we have identified several Star-PAP direct target genes and mRNAs by chromatin immunoprecipitation (ChIP) and RNA immunoprecipitation (RIP). One of these genes encodes the cytoprotective enzyme Heme Oxygenase-1 (HO-1). The HO-1 mRNA requires both PIPKIa and Star-PAP for its 3’ end processing and both enzymes interact with the HO-1 gene and mRNA as shown by ChIP and RIP approaches. HO-1 expression is highly inducible by the anti-oxidative response pathways. The anti-oxidant response pathways are required for the assembly of Star-PAP and PIPKIa into the RNA Pol II complex. The resulting Star-PAP complex when isolated from stimulated cells showed a ~40-fold increase in poly(A) polymerase activity. A conceptual model summarizing these results is depicted in Fig. 5. In this model Star-PAP or canonical PAP is selectively assembled into a RNA Pol II complex with components of the 3’ processing machinery. The oxidative stress-signaling pathway regulates both the assembly and activity of the Star-PAP complex. Further, the activation of Star-PAP by PIP2 is the first example of an mRNA modifying enzyme that is regulated by a small molecule second messenger. 

The biological significance of this pre-mRNA processing pathway will be a continuing emphasis by addressing the following questions. Does Star-PAP make a unique 3’ end tail on its target mRNAs? What are the signals for gene recognition by Star-PAP? Are there gene specific sequence elements required for Star-PAP assembly? Can genes use both Star-PAP and canonical PAP for 3’ processing under different conditions? What is the underlying mechanism for inducible Star-PAP assembly into the RNA Pol II complex? Are there Star-PAP targets that are not regulated by the anti-oxidative response? Do other signaling pathways modulate Star-PAP? What is the structural basis for Star-PAP activation by PIP2? Finally, the in vivo function of Star-PAP will be defined by generating Star-PAP null and transgenic mice.

PIPKIIb and PI5P regulate a nuclear ubiquitylation pathway. The type IIb PIP kinase (PIPKIIb) also targets to nuclear speckles. We identified a functional interaction between the PIPKIIβ and the speckle-type POZ domain protein (SPOP). Our data demonstrate that PIPKIIβ and SPOP interact and co-localize at nuclear speckles. SPOP is a scaffolding protein containing two protein-protein interaction domains: an N-terminal MATH (Meprin And TRAF Homology) domain, and a C-terminal POZ (POxvirus and Zinc finger) domain. Since SPOP only contains protein-protein interaction domains, Y2H screens were used to identify SPOP interacting proteins. From the Y2H screens we identified key members of the SUMOylation and ubiquitylation machinery, as well as several proteins that are substrates for SUMOylation and ubiquitylation. The SPOP interactors include Cul3 and RBX1 that play roles in ubiquitylation pathways. This is consistent, as SPOP is a member of the family of POZ and MATH domain containing proteins that act as substrate specificity factors for Cul3-based E3 ubiquitin ligase complexes. These proteins associate with Cul3 through their POZ domain and recruit substrates through a second protein-protein interaction domain. In the case of SPOP, its MATH domain could bind the substrate and recruit it to the E3 ligase for unbiquitination. Indeed, Daxx or death domain-associated protein 6 interacts with SPOP and immerging evidence in the literature demonstrates that SPOP is required for the ubiquitination of Daxx.The nuclear protein Daxx has been proposed to plays roles in apoptosis and transcriptional repression. SPOP interacts with an additional group of putative substrates that play roles in apoptosis and transcriptional regulation.

The interaction of PIPKIIb with the MATH domain suggested that it is a substrate for ubiquitylation. The primary sequence of PIPKIIb indicates that there are ubiquitylation consensus sites on the surface of the PIPKIIb structure, and we have shown that PIPKIIb is modified by ubiquitylation in a SPOP dependent manner. However, the interaction between SPOP and PIPKIIb may also be regulatory in nature. Recently it was shown that PIPKIIb is phosphorylated by p38 MAPK in response to genotoxic stress. Phosphorylation of PIPKIIb by p38-MAPK inhibited PI5P 4-kinase activity, resulting in accumulation of nuclear PI5P that putatively transduced the stress signal to promote p53 activation and apoptosis. We have shown that PIPKIIb unbiquitination via SPOP requires activation of p38-MAPK via the MKK6 pathway and is blocked by the p38-MAPK specific inhibitor SB203580.

Most significant, the ubiquitylation of SPOP substrates was stimulated several fold by expression of a kinase dead PIPKIIb, which increases nuclear PI5P by blocking the conversion of PI5P to PIP2. This raised the possibility that PI5P activates a SPOP ubiquitylation pathway. To explore this hypothesis, phosphatidylinositol-4,5-bisphosphate 4-phosphatase (PIP2 4-PPtase) was expressed and this stimulated the ubiquitylation of SPOP substrates Daxx and Pdx1 by ~50-fold. The enhanced of ubiquitylation was dependent upon SPOP, MKK6, and p38 as well as the activity of the PIP2 4-PPtase. The modulation of ubiquitylation pathways in the nucleus by PI5P is a unique and unexpected mechanism that appears to be a component of the oxidative and/or genotoxic stress pathways. Recently, we have shown that specific genotoxic stress pathways stimulate this pathway.

Signal transduction within the nucleus is an emerging field with significant implications for human health. We have discovered two novel phosphoinositide based signaling pathways that have changed the way we think about gene expression and the regulation of nuclear protein modification. The laboratory is now well positioned to fully define these pathways and expand upon their implications for human health.  This work is currently supported by NIH grant RO1 GM051968 and the Kellett endowment.

Research Scientists in the Anderson Laboratory

• Dr. Kun Ling, Ph.D., Scientist (Ph.D. Molecular Biology, Institute of Cell Biology, Shanghai)
• Dr. Christy Barlow, Ph.D., Postdoc (Ph.D. - Molecular Biology, University of Vermont)
• Dr. Yue Sun, Ph.D., Postdoc (Ph.D. - Molecular Biology, Institute of Cell Biology, Shanghai)

Graduate Students:

• Andrew Hedman (B.S. - University of Illinois)
• Matthew Bunce (B.S. - University of Virginia)
• David Mellman (B.S. - University of Colorado)
• Nicholas Schill (B.S. - St. Norbert College)
• Jessica Heck, (B.S. - Minnesota State University-Moorhead)

Associate Research Specialist:

Research Assistants:

• Kristin Mueller
• Lisa Christenson
• Lisa Goetter

Positions available:

• Postdoctoral and Graduate Student positions are available for 2007.

• Bunce MW, Boronenkov IV, and Anderson RA. (2008). Coordinated Activation the Nuclear Ubiquitin Ligase Cul3/Spop by the Generation of Phosphatidylinositol 5-Phosphate. J Biol Chem.  In press.

• Gonzales ML, Mellman DL, and Anderson RA. (2008). Star-PAP is Associated with and Phosphorylated by the Protein Kinase CKIa that is also Required for Expression of Select Star-PAP Target Messenger RNA. J Biol Chem.  In press.

• Bunce MW and Anderson RA. (2007). The nuclear speckle targeted type IIb PIP kinase modulates a stress dependent ubiquitination pathway. J Biol Chem.  In review.

• Mellman DL, Gonzales ML, Song C, Wang P, Kendziorski C, and Anderson RA. (2007). A Novel Phosphatidylinositol 4,5-Bisphosphate Regulated Nuclear Poly(A) Polymerase is Required for Efficient  3’-end Processing of Select mRNA. Nature. In review.

• El Sayegh TY, Arora PD, Laschinger C, Janmey PA, Anderson RA, McCulloch CA. (2007). Phosphatidylinositol-4,5 bisphosphate Produced by PIP5KIg Regulates Gelsolin, Actin Assembly and Adhesion Strength of N-Cadherin Junctions. Mol Biol Cell. In Press.
  
  
• Sun Y, Wagoner MP, Ling K, and Anderson RA. (2007). Type Ig PIP Kinase is Required for EGF-Stimulated Directional Cell Migration. J Cell Biol. In Press.
  
  
• Wagoner MP, Ling K, and Anderson RA.  (2007). Tracking the Transport of E-Cadherin to and from the Plasma Membrane.  Methods in Mol Biol. In press.
  
  
• Heck JN, Mellman DL, Ling K, Sun Y, Wagoner MP, Schill NJ, and Anderson RA. (2007). A conspicuous connection: structure defines function for the phosphatidylinositol-phosphate kinase family. Crit Rev. Biochem Mol Biol. 42:15-39. PDF PMID 17364683
  
  
• Ling K, Bairstow SF, Carbonara C, Turbin DA, Huntsman DG, and Anderson RA. (2007). Type Ig phosphatidylinositol phosphate kinase modulates adherens junction and E-cadherin trafficking via a direct interaction with m1B adaptin. J Cell Biol. 176:343-53. PDF PMID 17261850
  

View more Articles