Bram G. Lambrus Current Address blambru1@jhmi.edu Permanent Address 13 N Collington Ave. (405) 742-7961 1423 Hickory Court Baltimore, MD 21231 Stillwater, OK 74074

EDUCATION Johns Hopkins School of Medicine (2013-present) Ph.D. Candidate in the Biochemistry, Cellular and Molecular Biology (BCMB) program Training in the lab of Andrew Holland Honors, Awards, and Activities:

• NSF GRFP Fellowship • Kelly Award • Student Representative on the BCMB Policy Committee

University of Illinois at Urbana-Champaign (2008-2011) Bachelor of Science in Biochemistry Minor: Chemistry Cumulative GPA: 3.98/4.0, Summa Cum Laude GRE Score: Quantitative – 800 (94%), Verbal – 720 (98%) Honors, Awards, and Activities:

• Bronze Tablet Honors – awarded to top 3% of graduates in each college (highest UIUC honor) • Thomas O. Sidebottom Award for Most Outstanding Senior Thesis, “Molecular

Characterization of the Planarian Photoreceptors” • Merck Index Award for Outstanding Senior in Biochemistry • Highest Departmental Distinction • James Scholar Honors – for completion of the 4-year James Scholars Honors Program • Organic Chemistry Supplemental Instructor • Dean’s List – 6 semesters • Krannert Performing Arts Center Student Association – Volunteer Usher (2008-2011)

Washington University in St. Louis (2007-2008)

Major: Biochemistry Cumulative GPA: 3.95/4.0 Honors, Awards, and Activities:

• Dean’s List – 2 semesters • Residential College Council Representative

Stillwater High School (2004-2007) Cumulative GPA: 4.0/4.0 Honors, Awards, and Activities:

• National Merit Scholar • Oklahoma Academic All-State

PUBLICATIONS Lambrus BG, Uetake Y, Clutario KM, Daggubati V, Snyder M, Sluder G, Holland AJ. (2015) p53

protects against genome instability following centriole duplication failure. J. Cell Bio. 210(1), 63-77. Moyer TC, Clutario KM, Lambrus BG, Daggubati V, Holland AJ. (2015) STIL binding to Plk4

activates kinase activity to promote centriole duplication. Journal of Cell Biology. 209(6), 863-78.

Lambrus BG, Cochet-Escartin O, Collins EM, Newmark PA, Collins JJ. (2015) Tryptophan hydroxylase is required for eye pigmentation in planarian Schmidtea mediterranea. PLoS One, doi: 10.1371/journal.pone.0127074.

Wang JT, Smith J, Chen BC, Schmidt H, Rasoloson D, Paix A, Lambrus BG, Calidas D, Betzig E, Seydoux G. (2014) Regulation of RNA granule dynamics by phosphorylation of serine-rich, intrinsically disordered proteins in C. elegans. Elife, doi: 10.7554/eLife.04591.

Collins JJ, Wang B, Lambrus BG, Tharp M, Iyer H, Newmark PA. (2013) Neoblast-like somatic stem cells in the human parasite Schistosoma mansoni. Nature, 494, 476-479.

Collins JJ, Hou X, Romanova EV, Lambrus BG, Miller CM, Saberi A., Sweedler JV, Newmark PA. (2010) Genome-Wide Analyses Reveal a Role for Peptide Hormones in Planarian Germline Development. PLoS Biol 8, e1000509. PRESENTATIONS Lambrus B. (2015). Centrosome loss activates a protective p53-dependent cell cycle arrest. Presented at

the BCMB program retreat. Lambrus B. (2015). Characterizing centriole biogenesis with auxin-inducible degradation of Plk4.

Presented at the NCI at a Centrosome Meeting. Lambrus B. (2015). Characterizing centriole biogenesis with auxin-inducible degradation of Plk4.

Presented at the NIH at a Mitosis Meeting. Lambrus B. (2014). Characterizing centriole biogenesis with auxin-inducible degradation of Plk4.

ASCB. (poster) Lambrus, B. (2012). Tryptophan hydroxylase is required for proper eye function in planarians. Presented for the Integrated Science Shorts Seminar Series at Princeton University. Lambrus B, Collins JJ, Nelson M, Newmark PA. (2009). Characterization and functional analysis of neuropeptides in planarian Schmidtea mediterranea. Society for Neuroscience, UIUC. (poster)

1

Centrosome loss activates a protective p53-dependent cell cycle arrest Bramwell G. Lambrus Abstract Centrosome function has been difficult to study due to a lack of specific tools that allow persistent and reversible centrosome depletion. Here, we combine gene targeting with an auxin-inducible degradation system to achieve rapid, titratable, and reversible control of Polo-like kinase 4 (Plk4), a master regulator of centrosome biogenesis. Depletion of Plk4 led to a failure of centrosome duplication that, surprisingly, produced an irreversible cell cycle arrest within a few divisions. This arrest was not caused by known mechanisms, such as chromosome segregation errors, oxidative stress, prolonged mitosis, or DNA damage. This provides evidence for a novel surveillance pathway, which triggers growth arrest in cells with abnormal centrosome number. A genome-wide CRISPR/Cas9 knockout screen identified p53 and 53BP1 as two components of the surveillance pathway. Knockout of either gene allowed continued proliferation in the absence of centrosome duplication, resulting in a population of cells lacking centrosomes. These cells exhibit increased frequencies of mitotic errors, highlighting the importance of the centrosome surveillance pathway for protecting genome integrity. In summary, we generate a valuable tool for the study of centrosome biology, and uncover a novel surveillance mechanism that prevents cell growth following centrosome duplication failure.

2

Centrosome loss activates a protective p53-dependent cell cycle arrest Bramwell G. Lambrus Centrosomes are the major microtubule organizing centers of animal cells, and are comprised of pericentriolar material (PCM), recruited around a pair of microtubule-based structures called centrioles (1, 2). Centrioles act as the centrosome organizer, and thus their duplication controls centrosome number. Like DNA, centriole duplication takes place exactly once per cell cycle, during S-phase (3). This tightly controlled process ensures the generation of two centrosomes that form the poles of the mitotic spindle, a role proposed to be important for successful bipolar divisions. Errors in centriole duplication lead to abnormal centrosome number, which often result in chromosome segregation errors and the production of aneuploid progeny (4), highlighting the importance of precise centriole duplication for genome stability. Despite the importance of centrosome copy number control, the duplication process remains poorly understood. This is due in part to a shortage of tools with which to interrogate centriole assembly. Polo-like kinase 4 (Plk4) has been identified as a conserved, dose-dependent regulator of centriole duplication, yet its regulation and substrates remain largely unknown (5, 6). Thus, tools to manipulate Plk4 function have great potential to inform us of centriole biology. Traditional approaches such as RNA interference and gene knockout approaches have been used to disrupt Plk4 function, but these strategies are slow-acting and not readily reversible. An alternative approach is chemical inhibition of Plk4 kinase activity. However, this has been hampered by the difficulty of developing specific Plk4 inhibitors (7, 8). To address the challenge of studying Plk4 and centriole biology, we developed a chemical genetics approach to rapidly and reversibly control Plk4 protein abundance in cells, exploiting an auxin-inducible degradation system endogenous to plants (9). In plants, auxin promotes the binding of the F-box protein osTIR1 to proteins containing an auxin-inducible degron (AID). In the presence of auxin, osTIR1 recruits AID-containing proteins to the SCF ubiquitin ligase, where they are ubiquitinated and targeted for proteasomal degradation. Ectopic expression of osTIR1 in mammalian cells generates an SCFTIR1 complex and enables auxin-inducible destruction of AID-tagged transgenes (10). We aimed to apply this system to post-translationally control the abundance of endogenous Plk4 in human cells (Fig. 1A). We used sequential rounds of gene targeting to knock-in the AID onto the C-terminus of both endogenous Plk4 alleles in human RPE-1 cells. Each Plk4 allele was tagged at the C-terminus with an HA or 3x FLAG tag, to facilitate detection. Two Plk4AID-HA/AID-3xFLAG (hereafter referred to as Plk4AID/AID) clones were obtained, and behaved similarly in all assays, displaying normal centrosome copy number and cell cycle profiles. Next, osTIR1-9xMyc was stably expressed in Plk4AID/AID cells, to place the stability of Plk4AID under the control of exogenous auxin (Fig. 1A). To test the responsiveness of the system, auxin (IAA) was added to Plk4AID/AID cells at different time points, and Plk4 levels were assessed by immunoprecipitation with a FLAG antibody and detection by immunoblot. We observed rapid depletion of Plk4 upon addition of IAA, with Plk4 falling below the limit of detection within 10 minutes of treatment (Fig. 1B). We tested the functional effect of depletion by analyzing centriole duplication 24 hours after induction, and confirmed that induced Plk4 depletion results in centriole duplication failure (Fig. 1C). Washout of IAA allowed Plk4 to return to original levels within 3 hours (data not shown). Altogether, we have generated a fast-acting, inducible tool with which to study centriole and centrosome biology.

3

4

Surprisingly, when this system was applied to study how cells respond to chronic centrosome duplication failure, we observed a penetrant cell cycle arrest 48 hours after depletion of Plk4 (Fig. 1D). Analysis of centriole content revealed that centriole copy number decreased dramatically in the first 48 hours, as cells continue to divide in the absence of centriole duplication (Fig. 1E). After this time, centriole number remained constant, consistent with the observed growth arrest. To understand the cause of the growth arrest, we examined how centrosome loss affects cell division by filming Plk4AID/AID cells co-expressing RFP-tubulin, GFP-H2B, and the centrosome marker GFP-Cep63 (11). Although we observed a modest increase in chromosome missegregation errors, the frequency of these errors was too small to account for the growth arrest (Fig. 1F). Interestingly, the average time in mitosis increased significantly over time in cells with reduced centrosome number (Fig. 1G). It has previously been reported that prolonged mitosis duration in RPE1 cells triggers a durable G1-arrest (12). We asked whether this could account for the cell cycle arrest observed after centrosome duplication failure. By lineage-tracing cells over time in IAA, we analyzed the mitosis duration of individual cells, as well as the fate of their daughters (Fig. 1H). We found that the growth arrest could not be explained by prolonged mitosis duration, as many cells arrest despite completing mitosis well under the tolerated time window. Previous work has shown that DNA damage, Hippo pathway activation or excessive oxidative stress can cause a cell cycle arrest (13). However, centrosome loss did not lead to a detectable increase in DNA damage (measured by γH2A.X phosphorylation) or Hippo pathway activation (revealed by LATS2 or YAP phosphorylation) in Plk4AID/AID cells (Fig. 1I). Growth in low-oxygen conditions also did not affect the growth arrest (data not shown). Taken together, our results indicate that the G1 arrest following centrosome loss is not due to any of the previously described mechanisms that control cell cycle progression. Since centrosome loss can promote mitotic errors, we propose that a “centrosome surveillance pathway” acts to protect genomic stability by preventing growth in cells with abnormal centrosome numbers. We reasoned that such a pathway would be mediated by components which, when depleted, would prevent the growth arrest. To identify components of this centrosome surveillance pathway, we established a genome-wide CRISPR/Cas9 knock-out screen (14) to look for loss-of-function mutations that allow continued growth following centrosome loss. We stably expressed Cas9 in Plk4AID/AID cells, then transduced these cells with the a genome-wide single guide RNA (sgRNA) library (Fig. 2A), which results in site-specific DNA cleavage and frame-shift knockout mutations by low-fidelity repair. Samples were collected before and 42 days after inhibiting centrosome duplication. Deep sequencing of sgRNAs revealed a dramatic enrichment for guides targeting p53 and 53BP1 in the selected population (Fig. 2B). We used 3-5 individual sgRNAs to validate these top candidates and showed by clonogenic assay that independent knockout of p53 and 53BP1 allowed cells to proliferate despite centrosome loss (Fig. 2C-D). Strikingly, continued growth in the absence of p53 or 53BP1 resulted in a >90% acentrosomal cell population (Fig. 2E, data not shown). To examine the nature of mitosis in cells lacking centrosomes, we co-expressed RFP-tubulin, GFP-H2B, and GFP-Cep63 in p53-depleted Plk4AID/AID cells and filmed chronically treated cells. While cytokinesis failure occurs only rarely in untreated cells, acentrosomal cells failed cytokinesis in 20% of mitoses, and chromosome missegregation increased from 15% to 40% (Fig. 2F). These results illustrate the importance of a centrosome surveillance pathway in preventing progression into a genomically unstable acentrosomal state.

5

The accurate transmission of genetic material is fundamental for cell and organismal viability. Therefore, processes involved in replicating and transmitting the genome, such as DNA replication and spindle assembly, require robust quality control mechanisms. By generating a chemical genetic system that enables interrogation of centrosome biology, we have uncovered a novel quality control pathway for monitoring centrosome status in cells. This p53- and 53BP1-dependent surveillance pathway prevents growth in the presence of abnormal centrosome number, protecting cells from entering unstable acentrosomal mitoses. Future work will focus on identifying the mechanism by which centrosome loss triggers p53 and 53BP1 signaling in the centrosome surveillance pathway. As abnormal centrosome copy number is prevalent in cancer cells, we hope that understanding the fundamental pathways that protect against the development of aberrant centrosome status and genome instability will allow us to better design therapeutic strategies for the prevention and treatment of the disease.

6

References 1. P. Gönczy, Towards a molecular architecture of centriole assembly. Nat Rev Mol Cell Biol. 13,

425–435 (2012).

2. E. A. Nigg, J. W. Raff, Centrioles, Centrosomes, and Cilia in Health and Disease. Cell. 139, 663–678 (2009).

3. M.-F. B. Tsou, T. Stearns, Controlling centrosome number: licenses and blocks. Current Opinion in Cell Biology. 18, 74–78 (2006).

4. N. J. Ganem, S. A. Godinho, D. Pellman, A mechanism linking extra centrosomes to chromosomal instability. Nature. 460, 278–282 (2009).

5. M. Bettencourt-Dias et al., SAK/PLK4 is required for centriole duplication and flagella development. Curr. Biol. 15, 2199–2207 (2005).

6. R. Habedanck, Y.-D. Stierhof, C. J. Wilkinson, E. A. Nigg, The Polo kinase Plk4 functions in centriole duplication. Nat Cell Biol. 7, 1140–1146 (2005).

7. A. J. Holland, D. W. Cleveland, Polo-like Kinase 4 Inhibition:A Strategy for Cancer Therapy? Cancer Cell, 1–3 (2014).

8. J. M. Mason et al., Functional characterization of CFI-400945, a Polo-like kinase 4 inhibitor, as a potential anticancer agent. Cancer Cell. 26, 163–176 (2014).

9. K. Nishimura, T. Fukagawa, H. Takisawa, T. Kakimoto, M. Kanemaki, An auxin-based degron system for the rapid depletion of proteins in nonplant cells. Nat Meth. 6, 917–922 (2009).

10. A. J. Holland, D. Fachinetti, J. S. Han, D. W. Cleveland, Inducible, reversible system for the rapid and complete degradation of proteins in mammalian cells. Proceedings of the National Academy of Sciences. 109, E3350–E3357 (2012).

11. J.-H. Sir et al., A primary microcephaly protein complex forms a ring around parental centrioles. Nat Genet. 43, 1147–1153 (2011).

12. Y. Uetake, G. Sluder, Prolonged prometaphase blocks daughter cell proliferation despite normal completion of mitosis. Curr. Biol. 20, 1666–1671 (2010).

13. N. J. Ganem et al., Cytokinesis failure triggers hippo tumor suppressor pathway activation. Cell. 158, 833–848 (2014).

14. O. Shalem et al., Genome-scale CRISPR-Cas9 knockout screening in human cells. Science. 343, 84–87 (2014).

MICHELLE LEVINE 915 S Wolfe St., Unit 261 | Baltimore, MD 21231

Levine.michelle.s@gmail.com 305-790-2200

EDUCATION Johns Hopkins University, School of Medicine, Baltimore, MD August 2013 – Present PhD program: Biochemistry, Cellular, and Molecular Biology Boston College, Chestnut Hill, MA BA 2010 Major: Biology, Dean’s List, Sr. Thea Bowman AHANA Honor Roll Universidad Pablo de Olavide, Seville, Spain Spring Semester 2009 Course work in History of the Mediterranean World and Spain, Spanish Civilization and Culture, English-Spanish Translation, International Marketing, and Biochemistry RELEVANT WORK EXPERIENCE Johns Hopkins University, School of Medicine, Baltimore, MD March 2014 – Present PhD Student, Laboratory of Dr. Andrew Holland, PhD

• Determining whether and how extra centrosomes play a role in tumorigenesis Brigham and Women’s Hospital, Boston, MA Technical Research Assistant II September 2011 – May 2013 Technical Research Assistant I September 2010 – September 2011

• Investigated the capacity of the non-peptide thrombopoietin receptor agonist, eltrombopag, to expand the CD34+CD38- population in liquid culture containing a megakaryocyte cytokine cocktail

• Evaluated the actions of eltrombopag in stimulating megakaryopoiesis in bone marrow cells from patients with multiple myeloma

• Worked on project investigating the efficacy of a specific JAK2 inhibitor, compound #1781, in suppressing the proliferation of human cell line HEL and primary CD34+ cells isolated from Polycythemia vera patients without altering growth characteristics of normal CD34+ cells

• Studied the anti-inflammatory properties of Vitamin D (VD) in an in vitro, pro-inflammatory setting. Recorded the effects of VD on erythroid-specific signaling pathways and on TNFα-induced suppression of erythroid progenitors

• Helped identify the effect of translation initiation factor eIF4E on CCAAT enhancer binding protein alpha (C/EBPα) activity in nucleophosmin 1 (NPM1) haploinsufficient cells

Brigham and Women’s Hospital, Boston, MA October 2009 – April 2010 Student Intern

• Worked under the supervision of Arati Khanna-Gupta, PhD, to elucidate the roles of C/EBPα p42, C/EBPα p30 and NPM1 in Acute Myeloid Leukemia (AML) and Myelodysplastic Syndrome (MDS) using transient transfections and luciferase assays

PUBLICATIONS Levine MS, Bakker B, Boeckx B, Moyett J, Lu J, Vitre B, Spierings DC, Lansdorp PM, Cleveland DW, Lambrechts D, Foijer F and Holland AJ. Centrosome amplification is sufficient to promote spontaneous tumorigenesis in mammals. Developmental Cell. 2017 Feb 6;40(3):313-322.e5. doi: 10.1016/j.devcel.2016.12.022. (PMID: 28132847).

Jeong JY, Levine MS, Abayasekara N, Berliner N, Laubach J, Vanasse GJ. The non-peptide

thrombopoietin receptor agonist eltrombopag stimulates megakaryopoiesis in bone marrow cells from patients with relapsed multiple myeloma. Journal of Hematology Oncology. 2015 Apr 16;8(1):37. doi: 10.1186/s13045-

015-0136-2 (PMID: 25886818). Levine MS and Holland AJ. Plk4 Shapes Up. Structure, 2014 Aug 5; ;22(8):1071-3. doi: 10.1016/j.str.2014.07.004 (PMID: 25099950). Payne E, Virgilio M, Narla A, Sun H, Levine M, Paw BH, Berliner N, Look AT, Ebert B, Khanna-Gupta A. L-Leucine improves anemia and developmental defects associated with Diamond-Blackfan anemia and del(5q)MDS by activating the mTOR pathway. Blood, 2012 Sep 13; 120(11): 2214-24 (PMID: 22734070).

Khanna-Gupta, A, Abayasekara N, Levine M, Bolli N, Sun H, Silver M, Virgilio M, Halene S, Sportoletti P, Pandolfi PP, Berliner N. Upregulation of eIF4E in Nucleophosmin 1 (NPM1) haploinsufficient cells results in changes in CCAAT enhancer binding protein alpha (C/EBPa) activity: implications in MDS and AML. Journal of Biological Chemistry, 2012 Jul 31 (E-pub ahead of print, PMID: 22851180).

ABSTRACTS AND PRESENTATIONS

Michelle Levine, Bjorn Bakker, Bram Boeckx, Julia Moyett, James Lu, Benjamin Vitre, Diana C. Spierings, Peter M. Lansdorp, Don W. Cleveland, Diether Lambrechts, Floris Foijer and Andrew J. Holland. Determining the role of centrosome amplification in tumorigenesis. (American Society of Cell Biology Annual Meeting, 2016, Abstract, Poster, and Microsymposium talk).

Michelle Levine and Andrew Holland. Elucidating the Role of Centrosome Amplification in Tumorigenesis.

(American Society of Cell Biology/International Forum for Cell Biology Annual Meeting, 2014, Abstract and Poster).

Michelle Levine, BA, Jee-Yeong Jeong, PhD, Nancy Berliner, MD and K. Gary J. Vanasse, MD. 1, 25-Dihydroxyvitamin D Suppresses TNFα-Mediated NF-κB Activation in K562 and Human CD34+ Cells and Promotes Recovery of Erythroid Colony Formation in Human CD34+ Cells (American Society of Hematology, 53rd Annual Meeting, 2011, Abstract and Poster).

Nirmalee Abayasekara, MS, Michelle Levine, BA, Niccolo Bolli, MD, Hong Sun, MD, Matthew Silver, BS, Navid Nia, Paolo Sportoletti, MD, Pier Paolo Pandolfi, MD, PhD, Nancy Berliner, MD and Arati Khanna-Gupta, PhD. Upregulation of eIF4E in Nucleophosmin 1 (NPM1) Haploinsufficient Cells Alters CCAAT Enhancer Binding Protein Alpha (C/EBPα) Activity: Implications for MDS and AML (American Society of Hematology, 53rd Annual Meeting, 2011, Abstract and Poster).

Jee-Yeong Jeong, Michelle S. Levine, Bertal H. Aktas, Michael Chorev, Nirmalee Abayasekara,

Matthew Silver, Nancy Berliner, Jose A. Halperin and Gary J. Vanasse. A N,N’-diarylurea inhibitor of protein translation initiation selectively suppresses the proliferation of CD34+ cells isolated from Polycythemia Vera patients via activation of the HRI-eIF2α pathway. (European Hematology Association, 17th Congress meeting, 2012, Poster). AWARDS AND FELLOWSHIPS NIH Ruth L. Kirschstein NRSA (F31) Fellow January 2017 – present Hans Joaquim Prochaska Young Investigator’s Day Award May 2017 JHU Graduate Student Association Poster Prize – 2nd place May 2017 American Society of Cell Biology Travel Award December 2016 American Society of Cell Biology Poster Prize – 1st place December 2016

American Society of Cell Biology Travel Award December 2014

VOLUNTEER EXPERIENCE Project Bridge, Baltimore, MD October 2015 – Present Community outreach volunteer

• Conduct scientific experiments at different community events to engage and educate the public

Thread (formerly known as Incentive Mentoring Program), Baltimore, MD Oct. 2013 – Oct. 2014 Mentor

• Engaged weekly with an at-risk, underperforming high school student to help with school work

BWH Science Club, Boston, MA November 2012 – December 2012 Teacher

• Led science classes to 4th grade students at Tobin Elementary school Alzheimer’s Association, Boston, MA August 2012 – July 2013 Outreach and Education volunteer

• Taught community members about the disease and what the organization does to help

Association for Women in Science—Massachusetts Chapter July 2012 – July 2013 Events Committee member

• Assisted in planning events to bring together, educate, and facilitate networking among women in scientific fields

Horizons for Homeless, Boston, MA April 2011 – October 2011 Playspace Activities Leader

• Mentored and supervised children living in a homeless shelter Colegio Claret, Sevilla, Spain January 2009 – May 2009 English Teacher’s Assistant

• Developed lesson plans and instructed fifth graders learning English

Student Admissions Program, Boston College September 2006 – December 2008 Tour Guide

• Familiarized prospective students and their families with Boston College Greeter

• Welcomed and provided information to prospective students and their families Casa Nueva Vida, Jamaica Plain, MA September 2007 – May 2008

• Supervised and tutored 32 children living in the homeless shelter • Assisted with English as a second language and computer training courses for 16 mothers

SKILLS

• Fluent in Spanish • Accomplished flute and piccolo player

!!!

1"

Centrosome amplification promotes spontaneous tumorigenesis in mammals Michelle S. Levine Abstract Centrosomes are organelles that are responsible for organizing the bipolar spindle in mitosis. Centrosome copy number is usually tightly controlled so that each cell contains exactly two centrosomes upon entering mitosis. However, in cancers, cells often contain extra centrosomes (known as centrosome amplification)—a feature that is strongly correlated with high tumor grade and poor prognosis. These extra centrosomes have been proposed to promote chromosome segregation errors that drive aneuploidy and malignancy. However, the causative link between centrosome amplification and tumorigenesis has been previously unstudied due to the lack of tools to manipulate centrosome number alone. Here, we exploit our knowledge of the master regulator of centrosome biogenesis, Plk4, to create a mouse model in which levels of this protein can be modestly overexpressed to drive centrosome amplification. With this model, we can drive robust centrosome amplification and aneuploidy in diverse tissue types. Using a mouse model of intestinal neoplasia, we observe that centrosome amplification causes an increase in tumor initiation. Most importantly, we find for the first time that centrosome amplification is sufficient to promote spontaneous tumorigenesis in an animal model. This work provides an answer to a long-standing question in the field of centrosome biology and supports the targeting of cells with extra centrosomes in cancer therapy.

Michelle S. Levine

2

Centrosome amplification promotes spontaneous tumorigenesis in mammals

Each time a cell divides it must make a complete copy of its entire genome and then segregate this genome such that both daughter cells receive all the genetic information required for further growth and development. Key to faithful cell division is the presence of exactly two centrosomes during mitosis to build the bipolar spindle apparatus that segregates the chromosomes. Cells begin the cycle with a single centrosome that duplicates only once to ensure cells have two copies of this organelle when they divide (1). This faithful control of centrosome number is deregulated in a wide array of tumor types, resulting in the acquisition of extra copies of centrosomes (referred to as centrosome amplification) (2). Centrosome amplification causes mitotic errors in cultured cells that result in chromosome missegregation and chromosomal rearrangements that are frequently observed in human tumors (3, 4).

Although centrosome amplification is recognized as a hallmark of genomically unstable

cancer (2), the experimental tools to test whether supernumerary centrosomes drive tumorigenesis have been lacking. Furthermore, previous studies have been unable to separate a role of centrosome amplification from established oncogene or tumor suppressor pathways (5, 6). Thus, despite a large body of circumstantial evidence linking extra centrosomes to the development of cancer, it remains untested whether supernumerary centrosomes actively drive tumorigenesis or whether they are simply a byproduct of cellular transformation.

To address this long-standing question, I exploited our improved knowledge of the

control of centrosome biogenesis to generate an animal model in which extra centrosomes can be inducibly generated in vivo. Our experimental model builds on evidence from our lab and others that Polo-like kinase 4 (Plk4) regulates centrosome copy number (7-9). Thus, I generated a mouse model in which Plk4 levels can be modestly elevated to drive centrosome amplification in the absence of additional genetic defects (Figure 1A). Plk4 expression was under tight doxycycline control and consequently, this transgenic mouse is hereafter referred to as Plk4Dox.

To provide an initial characterization of our model, I derived mouse embryonic

fibroblasts (MEFs) and exposed them to doxycycline in vitro. This led to a very modest elevation in Plk4 levels (1.5 fold), but a robust increase in centrosome copy number, with ~70% of cells displaying centrosome amplification by three days after doxycycline addition (Figure 1B and 1C). As expected, these supernumerary centrosomes promoted chromosome missegregation leading to aneuploidy (Figure 1D).

To characterize the extent of Plk4 overexpression and centrosome amplification in vivo,

mice were sacrificed after 1 month of doxycycline treatment. Plk4 mRNA levels were increased in all tissues examined, with the exception of the brain where overexpression was not expected, due to the locus at which Plk4 is expressed (Figure 1E). Increased levels of Plk4 drove persistent centrosome amplification across multiple tissue types, validating this model as a tool to test the long-term effects of centrosome amplification in vivo (Figure 1F).

To examine the link between extra centrosomes and mitotic errors in vivo, we examined

the levels of aneuploidy in different tissue types. In the spleen, total chromosome numbers were quantified from mice treated with doxycycline for 1 month or 8 months. Centrosome amplification increased the number of aneuploid cells over time in the spleen (Figure 1G). I used whole genome sequencing of single cells to identify the karyotypes of adult skin cells. In control

Michelle S. Levine

3

animals, all cells sequenced were diploid. However, in mice with centrosome amplification, up to one-third of epidermal cells sequenced were aneuploid (Figure 1H). Therefore, centrosome amplification promotes chromosome missegregation and aneuploidy in vivo.

!!!!!!!!!!! !!!!!!! Figure 1. Modest Plk4 overexpression drives centrosome amplification and aneuploidy in vitro and in vivo. (A) Schematic of the mouse model created for doxycycline-inducible overexpression of Plk4. (B) Quantification of Plk4 levels at the centrosome in Plk4Dox mouse embryonic fibroblasts (MEFs). N=3, >150 centrosomes per experiment. (C) Quantification of the level of centrosome amplification in Plk4 Dox MEFs. N=3, >150 centrosomes per experiment. (D) Quantification of percent of cells containing <2 or >2 copies of chromosomes 15 or 16, as determined by FISH. N=3, >150 cells per experiment. (E) Graph showing Plk4 mRNA expression levels relative to controls in different tissues, as determined by qPCR analysis. N=3, performed in triplicate. (F) Quantification of centrosome amplification in control and Plk4Dox mice. N=3, >150 cells per experiment. (G) Graph showing quantification of severely aneuploid cells (4N ± >2 chromosomes) of splenocytes from 1-month or 8-month old control and Plk4Dox mice. N=3, >120 cells per experiment. (H) Table showing fraction of aneuploid cells determined by single-cell sequencing of control and Plk4Dox epidermal cells. All data represent the means ±SEM. *P < 0.05, **P < 0.01, ***P < 0.001; two-tailed Student’s t-test. Scale bars represent 10 µm. !

Michelle S. Levine

4

To establish whether centrosome amplification could affect the initiation and/or progression of tumors, I used a well-established mouse model of intestinal neoplasia (10). The power of this model is that tumor initiation and progression can be distinguished and quantified by measuring total polyp number and size, respectively. Using this model, I established that extra centrosomes promote increased initiation, but not progression, of intestinal tumors (Figure 2A and 2B).

To determine the long-term consequence of centrosome amplification, I aged a cohort of

control and Plk4Dox mice in a survival study. Strikingly, mice with centrosome amplification developed spontaneous tumors starting at about 9 months of age (Figure 2C). Animals with supernumerary centrosome developed spontaneous lymphomas, squamous cell carcinomas, and a sarcoma. These tumors contained robust centrosome amplification (Figure 2D).

Low-coverage whole genome sequencing revealed that the tumors that develop in mice

with centrosome amplification are highly aneuploid (Figure 2E). This supports the hypothesis that centrosome amplification drives chromosome missegregation and aneuploidy to promote tumorigenesis. Consistent with this proposal, lymphomas and squamous cell carcinomas exhibited recurrent aneuploidies, suggesting that particular karyotypes offer a selective advantage during tumor evolution.

Here, I have employed Plk4 overexpression as a tool to drive the production of extra

centrosomes. Although there is currently no strong evidence to support roles of Plk4 outside of centrosome biogenesis, I cannot rule out additional functions of this kinase. To address this, I transiently dosed mice for one month with doxycycline. While Plk4 levels return to baseline after one month, extra centrosomes persist in these mice over one year later (Figure 2F). This is consistent with the fact that extra centrosomes acts as templates for the subsequent duplication cycles and are stably propagated in the absence of continued increases in Plk4 levels. Importantly, animals that were transiently-dosed with doxycycline developed a similar spectrum of tumors as the chronically-treated mice—lymphomas, sarcomas and squamous cell carcinomas (Figure 2G). These data strongly suggest that extra centrosomes, and not increased levels of Plk4 per se, drive tumor formation in this model.

Centrosome number is normally tightly controlled in cells to maintain genomic integrity.

Cancer cells often contain too many centrosomes, but the contribution of these extra centrosomes to tumorigenesis has remained controversial. We show that centrosome amplification is sufficient to promote tumorigenesis by promoting chromosome segregation errors. This provides the first direct causative link between centrosome amplification and tumorigenesis.

Tumors that develop in mice with extra centrosomes recapitulate the hallmark karyotype

diversity that is frequently observed in human cancers. This suggests that our animal model may closely reflect the phenotypes of genetically unstable human tumors. Finally, our model offers an experimental platform to test the therapeutic targeting cells with extra centrosomes in cancer.

Michelle S. Levine

5

Figure 2. Centrosome amplification is sufficient to drive spontaneous tumorigenesis. (A and B) Graphs plotting total number of tumors (A) and average tumor size (B) in control (APCmin/+) and experimental (APCmin/+; Plk4Dox) mice treated with doxycycline from 1 week of age and sacrificed at 12 weeks old. Representative images of polyps in control and experimental mice (right). (C) Kaplan-Meier survival curve showing tumor-free survival of Plk4Dox (red line) and C57BL/6J control (blue line) mice chronically treated with doxycycline from 1-2 months of age. (D) Quantification of centrosome amplification in centrosome amplification-driven lymphomas (lymph.) and squamous cell carcinomas (SCC) as compared with p53-/--driven lymphomas. (E) GISTIC analysis of all lymphomas and squamous cell carcinomas from mice with centrosome amplification. Red bars represent gains of the respective chromosomes and length of bar represents the statistical significance of the recurrent gain of that chromosome. (F) Graph showing quantification of cells with centrosome amplification in mice transiently dosed with doxycycline. (G) Kaplan-Meier survival curve showing overall survival of mice overexpressing Plk4 for one month (red line) compared with controls (blue line). A dip in the curve represents a mouse that died in the study, and black shapes indicate mice that died of a tumor(s). All data represent the means ±SEM. *P < 0.05, **P < 0.01, ***P < 0.001; two-tailed Student’s t-test. Scale bars represent 10 µm. !!

Michelle S. Levine

6

References

1. E. A. Nigg, J. W. Raff, Centrioles, centrosomes, and cilia in health and disease. Cell 139, 663-678 (2009).

2. J. Y. Chan, A clinical overview of centrosome amplification in human cancers. International journal of biological sciences 7, 1122-1144 (2011).

3. N. J. Ganem, S. A. Godinho, D. Pellman, A mechanism linking extra centrosomes to chromosomal instability. Nature 460, 278-282 (2009).

4. W. T. Silkworth, I. K. Nardi, L. M. Scholl, D. Cimini, Multipolar spindle pole coalescence is a major source of kinetochore mis-attachment and chromosome mis-segregation in cancer cells. PLoS ONE 4, e6564 (2009).

5. P. A. Coelho et al., Over-expression of Plk4 induces centrosome amplification, loss of primary cilia and associated tissue hyperplasia in the mouse. Open Biol 5, 150209 (2015).

6. O. Sercin et al., Transient PLK4 overexpression accelerates tumorigenesis in p53-deficient epidermis. Nat Cell Biol 18, 100-110 (2016).

7. M. Bettencourt-Dias et al., SAK/PLK4 is required for centriole duplication and flagella development. Curr Biol 15, 2199-2207 (2005).

8. R. Habedanck, Y. D. Stierhof, C. J. Wilkinson, E. A. Nigg, The Polo kinase Plk4 functions in centriole duplication. Nat Cell Biol 7, 1140-1146 (2005).

9. A. J. Holland, W. Lan, D. W. Cleveland, Centriole duplication: A lesson in self-control. Cell Cycle 9, 2731-2736 (2010).

10. L. K. Su et al., Multiple intestinal neoplasia caused by a mutation in the murine homolog of the APC gene. Science 256, 668-670 (1992).

Kyle Scott Severson, B.S. The Johns Hopkins University School of Medicine The Solomon H. Snyder Department of Neuroscience Brain Science Institute 855 N Wolfe St, Room 287 Baltimore, Maryland 21205 Phone: 608-469-9464 Email: ksevers2@jhmi.edu EDUCATION2012-present Ph.D., Neuroscience, expected August, 2018

The Johns Hopkins University, Baltimore, MD Thesis: “Primary afferent encoding of whisker self-motion and active touch” Advisor: Prof. Daniel H. O’Connor Committee members: Xinzhong Dong, Jeremy Nathans, and Jeremiah Cohen

2008-2012 B.S., Biochemistry, Biology

University of Wisconsin, Madison, WI

OTHER TRAINING 2011-2012 Undergraduate researcher, Center for Sleep and Consciousness, University of

Wisconsin, Madison, WI Research advisor: Prof. Chiara Cirelli 2008-2011 Undergraduate researcher, Department of Biochemistry, University of

Wisconsin, Madison, WI Research advisor: Prof. Hector F. DeLuca

PUBLICATIONS

1. Severson KS*, Xu D*, Van de Loo M, Bai L, Ginty DD, O’Connor DH. Active touch and self-motion encoding by Merkel cell-associated afferents. Neuron. 2017. 94, 666–676.

2. Yang H*, Kwon SE*, Severson KS, O'Connor DH. Origins of choice-related activity in mouse somatosensory cortex. Nat Neurosci. 2016. 19(1): 127-34.

3. Wang Y, Marling SJ, Beaver EF, Severson KS, Deluca HF. UV light selectively inhibits spinal cord inflammation and demyelination in experimental autoimmune encephalo-myelitis. Arch Biochem Biophys. 2015. 567: 75-82.

4. Wang Y, Marling SJ, McKnight SM, Danielson AL, Severson KS, Deluca HF. Suppression of experimental autoimmune encephalomyelitis by 300-315nm ultraviolet light. Arch Biochem Biophys. 2013. 536(1): 81-6.

5. Wang Y, Marling SJ, Zhu JG, Severson KS, DeLuca HF. Development of experimental autoimmune encephalomyelitis (EAE) in mice requires vitamin D and the vitamin D receptor. Proc Natl Acad Sci U S A. 2012. 109(22): 8501-4.

6. Becklund BR, Severson KS, Vang SV, DeLuca HF. Ultraviolet radiation suppresses experimental autoimmune encephalomyelitis independent of vitamin D production. Proc Natl Acad of Sci USA. 2010. 107(14): 6418-23. *Co-first authors

OTHER PUBLICATIONS

1. Severson KS, O'Connor DH. Active Sensing: The rat's nose dances in step with whiskers, head, and breath. Curr Biol. 2017, 27(5): R183-R185.

PRESENTATIONS

1. Severson KS. Primary afferent encoding of whisker self-motion. Oral presentation at: Department of Neuroscience Annual Retreat. September, 2017. Cambridge, MD.

2. Severson KS, Xu D, Bai L, Ginty DD, O’Connor DH. Touch and self-motion encoding by Merkel-cell associated primary afferents. Poster presented at: Neuroscience 2016. November, 2016. San Diego, CA.

3. Severson KS, Xu D, O’Connor DH. Stimulus feature coding by trigeminal ganglion neurons during natural whisking against an object. Poster presented at: Neuroscience 2015. November, 2015. Chicago, IL.

TECHNIQUES

• In vivo electrophysiology in head-fixed, behaving mice • High-speed video measurement of mechanics and kinematics • Analysis of large datasets in MATLAB • Optogenetic identification • Mouse husbandry and breeding • Fluorescence microscopy

TEACHING EXPERIENCE 2014-2016 Undergraduate mentor, Johns Hopkins University Fall 2013 Teaching assistant, Neuroscience and Cognition course PROFESSIONAL MEMBERSHIPS 2015-present Society for Neuroscience

Mechanical encoding during active touch and self-motion Kyle S. Severson, B.S. The Solomon H. Snyder Department of Neuroscience, Brain Science Institute, and The Johns Hopkins University School of Medicine Touch perception depends on integrating signals from multiple types of peripheral mechanoreceptors. Merkel-cell associated afferents are thought to play a major role in form perception by encoding surface features of touched objects. However, Merkel afferent activity has not been directly measured during active touch. Merkel and unidentified slowly adapting afferents in the whisker system of behaving mice respond robustly during active touch and, surprisingly, respond during self-motion. Touch responses were dominated by sensitivity to bending moment (torque) at the base of the whisker and its rate of change. Self-motion responses encoded whisk phase, the position within a whisk cycle, not absolute whisker angle. These phase-tuned responses arose from stresses reflecting whisker inertia and activity of specific muscles. Thus, Merkel afferents send to the brain multiplexed information about surface features and whisker position, suggesting that touch and proprioception converge at the earliest neural level.

Touch perception originates in mechanosensory endings in the skin. Distinct types of mechanosensory neurons, distinguished by the structure of their nerve endings, have been described in many mammalian systems, including the primate finger tip1 and mouse hairy skin2. Merkel cell-associated primary afferents are thought to be important for the perception of spatial form of touched objects and are concentrated in areas with high sensitivity, such as human fingertips

and rodent whiskers. Whereas we humans use our hands to explore the tactile world, mice and rats typically sweep their whiskers to sense their physical surroundings. Mouse whiskers serve as a powerful model for us to understand how the brain encodes and integrates sensorimotor information because of well-mapped circuits, control of touch input, and genetic accessibility. It is also relatively straightforward to mechanically model the whisker compared to the human hand. For these reasons, we developed a preparation to study how Merkel afferents encode mechanics during active whisking.

We recorded extracellularly from primary afferent neurons with single whisker receptive fields. Merkel afferents were identified using an optogenetic tagging technique in the TrkCCreER mouse line expressing channelrhodopsin specifically in Merkel afferents3. We captured high speed video of this whisker and recorded afferent responses as the mouse whisked against a thin pole. We then aimed to correlate contact forces with resultant spiking and determine which mechanical features best explain the spiking activity of identified Merkel afferents.

When the whisker contacts an object, it bends and causes changes in whisker curvature measured in the video. We input this curvature and post hoc measurements of the whisker into established models of whisker mechanics4. The whisker model accurately estimates force (�⃑�) exerted by the stationary pole and torque, or moment (M0), exerted at the base of the whisker for each independent video frame (Figure 1A). M0 is expected to cause spiking in mechanosensory afferents with endings in the whisker follicle5,6. Here, we found that Merkel and unidentified afferents responded primarily during one direction of contact, either protraction (pushing forward) or retraction (pulling back). M0 is positive during protraction and negative

Figure 1. Mechanical modeling predicts spiking in SA and Merkel afferents. (A) Video frame of whisker showing normal force (�⃑�) and bending moment (�⃑⃑⃑�0) exerted by pole contact. (B) Moment was inputted to a simple viscoelastic model. Total stress (σtotal) resulted from the sum of stress on a spring (σspring) and a dashpot (σdashpot). (C) The viscoelastic model predicted spikes for slowly adapting (SA, black) and Merkel (blue) units as well as a generalized additive model (GAM, mean ± 95% CI). (D) Moment (top), modeled stresses (middle), and spiking (grey ticks, actual spike times; black, predicted spike rate) during an example whisker touch. Modified from Severson et al. 2017. during retraction. We applied a generalized additive model (GAM) to determine which combination of mechanical variables best explain spiking. Briefly, M0 and its rate of change, M0’, predicted spiking the best. Interestingly, sensitivity to these two variables seemed to interact in a more or less linear fashion. It is straightforward to understand why

M0 is related to spiking: bending at the base causes a deformation, or strain, on the follicle and results in stress on mechanosensory neurons. However, why M0’ is important was, at first, less clear. We adopted a theoretical framework that models the whisker follicle as a viscoelastic medium. In this Kelvin-Voigt model, M0 causes strain in elastic spring element and a viscous dashpot element placed in parallel (Figure 1B). Elastic stress in the spring is proportional to strain (i.e. M0). Viscous stress in the damper is proportional to the rate of change of strain (i.e. M0’). As an example, resistance in a fluid-filled syringe gets stronger the faster you push it. In this parallel arrangement, stresses in the spring and dashpot sum linearly to give total stress (Figure 1B,D). Our predicted spike rate scales linearly with the total mechanical stress (Figure 1D). The mechanical model predicted spikes as well as the GAM (Figure 1C, r = 0.66 vs. 0.69, respectively) with fewer free parameters (5 for mechanical model). The great performance of this simple model suggests that Merkel and unidentified afferent spiking is related to M0 via sensitivity to mechanical stress.

Merkel afferents are described as slowly adapting, due to their ability to sustain spiking during a sustained stimulus. Our framework predicts that elastic coupling of whiskers to Merkel endings in the follicle is necessary for their slowly adapting characteristics. In contrast, rapidly adapting afferents spike during the dynamic portion of a stimulus, but cease spiking during a sustained stimulus. Viscoelasticity is thought to underlie the rapid adaptation of the Pacinian corpuscle in mammals7. The Pacinian’s unique ability to encode high frequency vibrations is attributed to specialized glial sheaths that encapsulate the nerve ending in several layers of viscous tissue, like layers of an onion. By virtue of these examples, mechanical filtering is a plausible general mechanism by which primary afferents encode a diverse array of features during touch.

Many of the same afferents (n = 13/33 total, n = 5/14 Merkel afferents) with touch responses also spiked during self-motion, in the absence of touch. Which mechanical variables explain these self-motion responses? We tracked the whisker position by measuring whisker angle (θ) near the base. From θ, we calculated kinematics, such as its derivatives, velocity (ω) and acceleration (α), and quantities in the frequency domain, such as amplitude and phase (ϕ). Amplitude describes the half-width of a whisker cycle, in degrees, and phase represents the position within a whisk cycle, measured in radians. In previous studies, phase coding was observed in whisker afferents5,8, brainstem circuits9,10, and whisker primary somatosensory cortex11. Phase coding in cortex depends on afferent input, as phase modulation disappears after infraorbital nerve cut12. The mechanism underlying primary afferent encoding of phase is unknown.

One prominent hypothesis is that phase coding is due to inertia13. Moment is exerted at the axis of rotation proportional to the object’s moment of inertia, I, and angular acceleration, α (Figure 2A). This is a version of Newton’s second law, 𝐹 = 𝑚 · 𝑎, but applied to a rotational plane. Assuming that spike rates increase monotonically with M0, this hypothesis makes two predictions: (1) spike rates scale with increasing acceleration and (2) spike rates scale with changes in I. We devised an experiment to test this hypothesis directly (Figure 2A). The intact whisker has a particular moment of inertia, depending on its radius and length. If a segment of the whisker is removed by cutting, its moment of inertia will be reduced. We recorded SA afferent spiking during whisking in air while the whisker was intact and after progressive cutting of distal segments of the whisker. With the whisker intact, several units appeared to have strong acceleration tuning,

Figure 2. (A) Inertial moment (�⃑⃑⃑�) is proportional to moment of inertia (I) and angular acceleration (�⃑�). (B) Example unit tuning curves for whisk phase (left) and acceleration (right, ± SEM) for each cutting condition and control handling (“sham”, colors, as in A). (C) Example phase tuning for fully cut condition (black, normalized) with overlaid with average normalized EMG traces for intrinsic protractor (IP, red) or extrinsic retractor (ER, orange). Pearson’s r values compare phase tuning and best EMG traces. (D) Population histogram of preferred phases for all intact whisker recordings (n = 28). Colored traces illustrate average normalized acceleration (+α, -α) and EMG traces (IP, ER) given phase. Modified from Severson et al. 2017.

preferring either positive or negative acceleration (Figure 2B). Spike rates increased with greater acceleration in the preferred direction. Acceleration tuning was unaffected after “sham” handling the whisker with forceps to simulate the cutting procedure. After cutting and decreasing the whisker’s moment of inertia near zero, acceleration tuning was abolished for several units (n = 6/13 units). Furthermore, preferred phase in these units aligned with the peak of either negative acceleration (ϕ = -π/4) or positive acceleration (ϕ = -3π/4), and phase tuning was abolished as well (Figure 2B). Self-motion responses that persisted (n = 7/13 units) after cutting were correlated with whisker velocity rather than acceleration. The phasic activation of whisking muscles is correlated with velocity14, so we speculate that self-motion responses after cutting can be explained by torque exerted internally by activation of specific muscles. Mouse whiskers are actuated by two groups of fast twitch muscles15: one intrinsic protractor (IP) and two extrinsic retractors (ER). Electromyogram (EMG) recordings in either or both of these muscle groups closely resembled phase tuning in recordings in which the whisker was fully cut (Figure 2C). The inertia hypothesis was partially correct. The first prediction was generally true because we observed acceleration tuning in a large portion of units (Figure 2D). Whisker cutting reduced spike rates in several units. Therefore, based on the results of these experiments and our mechanical framework, we propose that stresses related to inertial moment and activation of whisking muscles contribute to self-motion responses in primary afferents. Mechanical input to primary afferents likely underlies phase tuning observed in the whisker system. Phase coding may serve as an important proprioceptive signal during whisker exploration. References:

1. Johnson, K.O., and Hsiao, S.S. (1992). Neural mechanisms of tactual form and texture perception. Annu Rev Neurosci 15, 227-250.

2. Abraira, V.E., and Ginty, D.D. (2013). The sensory neurons of touch. Neuron 79, 618-639.

3. Bai, L., Lehnert, B.P., Liu, J., Neubarth, N.L., Dickendesher, T.L., Nwe, P.H., Cassidy, C., Woodbury, C.J., and Ginty, D.D. (2015). Genetic identification of an expansive mechanoreceptor sensitive to skin stroking. Cell 163, 1783-1795.

4. Pammer, L., O'Connor, D.H., Hires, S.A., Clack, N.G., Huber, D., Myers, E.W., and Svoboda, K. (2013). The mechanical variables underlying object localization along the axis of the whisker. J Neurosci 33, 6726-6741.

5. Campagner, D., Evans, M.H., Bale, M.R., Erskine, A., and Petersen, R.S. (2016). Prediction of primary somatosensory neuron activity during active tactile exploration. Elife 5.

6. Bush, N.E., Schroeder, C.L., Hobbs, J.A., Yang, A.E., Huet, L.A., Solla, S.A., and Hartmann, M.J. (2016). Decoupling kinematics and mechanics reveals coding properties of trigeminal ganglion neurons in the rat vibrissal system. Elife 5.

7. Fleming M.S., Luo W. (2013). The anatomy, function, and development of mammalian Aβ low-threshold mechanoreceptors. Front Biol (Beijing). 1;8(4).

8. Khatri, V., Bermejo, R., Brumberg, J.C., Keller, A., and Zeigler, H.P. (2009). Whisking in air: encoding of kinematics by trigeminal ganglion neurons in awake rats. J Neurophysiol 101, 1836-1846.

9. Moore, J.D., Mercer Lindsay, N., Deschenes, M., and Kleinfeld, D. (2015). Vibrissa self-motion and touch are reliably encoded along the same somatosensory pathway from brainstem through thalamus. PLoS Biol 13, e1002253.

10. Wallach, A., Bagdasarian, K., and Ahissar, E. (2016). On-going computation of whisking phase by mechanoreceptors. Nat Neurosci 19, 487-493.

11. Curtis, J.C., and Kleinfeld, D. (2009). Phase-to-rate transformations encode touch in cortical neurons of a scanning sensorimotor system. Nat Neurosci 12, 492-501.

12. Poulet J.F.A and Petersen C.C.H. (2008) Internal brain state regulates membrane potential synchrony in barrel cortex of behaving mice. Nature 454, 881-885.

13. Quist B.W., Seghete V., Huet L.A., Murphey T.D., Hartmann M.J. (2014) Modeling forces and moments at the base of a rat vibrissa during noncontact whisking and whisking against an object. J Neurosci 23;34(30):9828-44.

14. Hill, D.N., Bermejo, R., Zeigler, H.P., and Kleinfeld, D. (2008). Biomechanics of the vibrissa motor plant in rat: rhythmic whisking consists of triphasic neuromuscular activity. J Neurosci 28, 3438-3455.

15. Dorfl, J. (1985). The innervation of the mystacial region of the white mouse: A topographical study. J Anat 142, 173-184.

Kyle Scott Severson, B.S. The Johns Hopkins University School of Medicine The Solomon H. Snyder Department of Neuroscience Brain Science Institute 855 N Wolfe St, Room 287 Baltimore, Maryland 21205 Phone: 608-469-9464 Email: ksevers2@jhmi.edu EDUCATION2012-present Ph.D., Neuroscience, expected August, 2018

The Johns Hopkins University, Baltimore, MD Thesis: “Primary afferent encoding of whisker self-motion and active touch” Advisor: Prof. Daniel H. O’Connor Committee members: Xinzhong Dong, Jeremy Nathans, and Jeremiah Cohen

2008-2012 B.S., Biochemistry, Biology

University of Wisconsin, Madison, WI

OTHER TRAINING 2011-2012 Undergraduate researcher, Center for Sleep and Consciousness, University of

Wisconsin, Madison, WI Research advisor: Prof. Chiara Cirelli 2008-2011 Undergraduate researcher, Department of Biochemistry, University of

Wisconsin, Madison, WI Research advisor: Prof. Hector F. DeLuca

PUBLICATIONS

1. Severson KS*, Xu D*, Van de Loo M, Bai L, Ginty DD, O’Connor DH. Active touch and self-motion encoding by Merkel cell-associated afferents. Neuron. 2017. 94, 666–676.

2. Yang H*, Kwon SE*, Severson KS, O'Connor DH. Origins of choice-related activity in mouse somatosensory cortex. Nat Neurosci. 2016. 19(1): 127-34.

3. Wang Y, Marling SJ, Beaver EF, Severson KS, Deluca HF. UV light selectively inhibits spinal cord inflammation and demyelination in experimental autoimmune encephalo-myelitis. Arch Biochem Biophys. 2015. 567: 75-82.

4. Wang Y, Marling SJ, McKnight SM, Danielson AL, Severson KS, Deluca HF. Suppression of experimental autoimmune encephalomyelitis by 300-315nm ultraviolet light. Arch Biochem Biophys. 2013. 536(1): 81-6.

5. Wang Y, Marling SJ, Zhu JG, Severson KS, DeLuca HF. Development of experimental autoimmune encephalomyelitis (EAE) in mice requires vitamin D and the vitamin D receptor. Proc Natl Acad Sci U S A. 2012. 109(22): 8501-4.

6. Becklund BR, Severson KS, Vang SV, DeLuca HF. Ultraviolet radiation suppresses experimental autoimmune encephalomyelitis independent of vitamin D production. Proc Natl Acad of Sci USA. 2010. 107(14): 6418-23. *Co-first authors

OTHER PUBLICATIONS

1. Severson KS, O'Connor DH. Active Sensing: The rat's nose dances in step with whiskers, head, and breath. Curr Biol. 2017, 27(5): R183-R185.

PRESENTATIONS

1. Severson KS. Primary afferent encoding of whisker self-motion. Oral presentation at: Department of Neuroscience Annual Retreat. September, 2017. Cambridge, MD.

2. Severson KS, Xu D, Bai L, Ginty DD, O’Connor DH. Touch and self-motion encoding by Merkel-cell associated primary afferents. Poster presented at: Neuroscience 2016. November, 2016. San Diego, CA.

3. Severson KS, Xu D, O’Connor DH. Stimulus feature coding by trigeminal ganglion neurons during natural whisking against an object. Poster presented at: Neuroscience 2015. November, 2015. Chicago, IL.

TECHNIQUES

• In vivo electrophysiology in head-fixed, behaving mice • High-speed video measurement of mechanics and kinematics • Analysis of large datasets in MATLAB • Optogenetic identification • Mouse husbandry and breeding • Fluorescence microscopy

TEACHING EXPERIENCE 2014-2016 Undergraduate mentor, Johns Hopkins University Fall 2013 Teaching assistant, Neuroscience and Cognition course PROFESSIONAL MEMBERSHIPS 2015-present Society for Neuroscience

Mechanical encoding during active touch and self-motion Kyle S. Severson, B.S. The Solomon H. Snyder Department of Neuroscience, Brain Science Institute, and The Johns Hopkins University School of Medicine Touch perception depends on integrating signals from multiple types of peripheral mechanoreceptors. Merkel-cell associated afferents are thought to play a major role in form perception by encoding surface features of touched objects. However, Merkel afferent activity has not been directly measured during active touch. Merkel and unidentified slowly adapting afferents in the whisker system of behaving mice respond robustly during active touch and, surprisingly, respond during self-motion. Touch responses were dominated by sensitivity to bending moment (torque) at the base of the whisker and its rate of change. Self-motion responses encoded whisk phase, the position within a whisk cycle, not absolute whisker angle. These phase-tuned responses arose from stresses reflecting whisker inertia and activity of specific muscles. Thus, Merkel afferents send to the brain multiplexed information about surface features and whisker position, suggesting that touch and proprioception converge at the earliest neural level.

Touch perception originates in mechanosensory endings in the skin. Distinct types of mechanosensory neurons, distinguished by the structure of their nerve endings, have been described in many mammalian systems, including the primate finger tip1 and mouse hairy skin2. Merkel cell-associated primary afferents are thought to be important for the perception of spatial form of touched objects and are concentrated in areas with high sensitivity, such as human fingertips

and rodent whiskers. Whereas we humans use our hands to explore the tactile world, mice and rats typically sweep their whiskers to sense their physical surroundings. Mouse whiskers serve as a powerful model for us to understand how the brain encodes and integrates sensorimotor information because of well-mapped circuits, control of touch input, and genetic accessibility. It is also relatively straightforward to mechanically model the whisker compared to the human hand. For these reasons, we developed a preparation to study how Merkel afferents encode mechanics during active whisking.

We recorded extracellularly from primary afferent neurons with single whisker receptive fields. Merkel afferents were identified using an optogenetic tagging technique in the TrkCCreER mouse line expressing channelrhodopsin specifically in Merkel afferents3. We captured high speed video of this whisker and recorded afferent responses as the mouse whisked against a thin pole. We then aimed to correlate contact forces with resultant spiking and determine which mechanical features best explain the spiking activity of identified Merkel afferents.

When the whisker contacts an object, it bends and causes changes in whisker curvature measured in the video. We input this curvature and post hoc measurements of the whisker into established models of whisker mechanics4. The whisker model accurately estimates force (�⃑�) exerted by the stationary pole and torque, or moment (M0), exerted at the base of the whisker for each independent video frame (Figure 1A). M0 is expected to cause spiking in mechanosensory afferents with endings in the whisker follicle5,6. Here, we found that Merkel and unidentified afferents responded primarily during one direction of contact, either protraction (pushing forward) or retraction (pulling back). M0 is positive during protraction and negative

Figure 1. Mechanical modeling predicts spiking in SA and Merkel afferents. (A) Video frame of whisker showing normal force (�⃑�) and bending moment (�⃑⃑⃑�0) exerted by pole contact. (B) Moment was inputted to a simple viscoelastic model. Total stress (σtotal) resulted from the sum of stress on a spring (σspring) and a dashpot (σdashpot). (C) The viscoelastic model predicted spikes for slowly adapting (SA, black) and Merkel (blue) units as well as a generalized additive model (GAM, mean ± 95% CI). (D) Moment (top), modeled stresses (middle), and spiking (grey ticks, actual spike times; black, predicted spike rate) during an example whisker touch. Modified from Severson et al. 2017. during retraction. We applied a generalized additive model (GAM) to determine which combination of mechanical variables best explain spiking. Briefly, M0 and its rate of change, M0’, predicted spiking the best. Interestingly, sensitivity to these two variables seemed to interact in a more or less linear fashion. It is straightforward to understand why

M0 is related to spiking: bending at the base causes a deformation, or strain, on the follicle and results in stress on mechanosensory neurons. However, why M0’ is important was, at first, less clear. We adopted a theoretical framework that models the whisker follicle as a viscoelastic medium. In this Kelvin-Voigt model, M0 causes strain in elastic spring element and a viscous dashpot element placed in parallel (Figure 1B). Elastic stress in the spring is proportional to strain (i.e. M0). Viscous stress in the damper is proportional to the rate of change of strain (i.e. M0’). As an example, resistance in a fluid-filled syringe gets stronger the faster you push it. In this parallel arrangement, stresses in the spring and dashpot sum linearly to give total stress (Figure 1B,D). Our predicted spike rate scales linearly with the total mechanical stress (Figure 1D). The mechanical model predicted spikes as well as the GAM (Figure 1C, r = 0.66 vs. 0.69, respectively) with fewer free parameters (5 for mechanical model). The great performance of this simple model suggests that Merkel and unidentified afferent spiking is related to M0 via sensitivity to mechanical stress.

Merkel afferents are described as slowly adapting, due to their ability to sustain spiking during a sustained stimulus. Our framework predicts that elastic coupling of whiskers to Merkel endings in the follicle is necessary for their slowly adapting characteristics. In contrast, rapidly adapting afferents spike during the dynamic portion of a stimulus, but cease spiking during a sustained stimulus. Viscoelasticity is thought to underlie the rapid adaptation of the Pacinian corpuscle in mammals7. The Pacinian’s unique ability to encode high frequency vibrations is attributed to specialized glial sheaths that encapsulate the nerve ending in several layers of viscous tissue, like layers of an onion. By virtue of these examples, mechanical filtering is a plausible general mechanism by which primary afferents encode a diverse array of features during touch.

Many of the same afferents (n = 13/33 total, n = 5/14 Merkel afferents) with touch responses also spiked during self-motion, in the absence of touch. Which mechanical variables explain these self-motion responses? We tracked the whisker position by measuring whisker angle (θ) near the base. From θ, we calculated kinematics, such as its derivatives, velocity (ω) and acceleration (α), and quantities in the frequency domain, such as amplitude and phase (ϕ). Amplitude describes the half-width of a whisker cycle, in degrees, and phase represents the position within a whisk cycle, measured in radians. In previous studies, phase coding was observed in whisker afferents5,8, brainstem circuits9,10, and whisker primary somatosensory cortex11. Phase coding in cortex depends on afferent input, as phase modulation disappears after infraorbital nerve cut12. The mechanism underlying primary afferent encoding of phase is unknown.

One prominent hypothesis is that phase coding is due to inertia13. Moment is exerted at the axis of rotation proportional to the object’s moment of inertia, I, and angular acceleration, α (Figure 2A). This is a version of Newton’s second law, 𝐹 = 𝑚 · 𝑎, but applied to a rotational plane. Assuming that spike rates increase monotonically with M0, this hypothesis makes two predictions: (1) spike rates scale with increasing acceleration and (2) spike rates scale with changes in I. We devised an experiment to test this hypothesis directly (Figure 2A). The intact whisker has a particular moment of inertia, depending on its radius and length. If a segment of the whisker is removed by cutting, its moment of inertia will be reduced. We recorded SA afferent spiking during whisking in air while the whisker was intact and after progressive cutting of distal segments of the whisker. With the whisker intact, several units appeared to have strong acceleration tuning,

Figure 2. (A) Inertial moment (�⃑⃑⃑�) is proportional to moment of inertia (I) and angular acceleration (�⃑�). (B) Example unit tuning curves for whisk phase (left) and acceleration (right, ± SEM) for each cutting condition and control handling (“sham”, colors, as in A). (C) Example phase tuning for fully cut condition (black, normalized) with overlaid with average normalized EMG traces for intrinsic protractor (IP, red) or extrinsic retractor (ER, orange). Pearson’s r values compare phase tuning and best EMG traces. (D) Population histogram of preferred phases for all intact whisker recordings (n = 28). Colored traces illustrate average normalized acceleration (+α, -α) and EMG traces (IP, ER) given phase. Modified from Severson et al. 2017.

preferring either positive or negative acceleration (Figure 2B). Spike rates increased with greater acceleration in the preferred direction. Acceleration tuning was unaffected after “sham” handling the whisker with forceps to simulate the cutting procedure. After cutting and decreasing the whisker’s moment of inertia near zero, acceleration tuning was abolished for several units (n = 6/13 units). Furthermore, preferred phase in these units aligned with the peak of either negative acceleration (ϕ = -π/4) or positive acceleration (ϕ = -3π/4), and phase tuning was abolished as well (Figure 2B). Self-motion responses that persisted (n = 7/13 units) after cutting were correlated with whisker velocity rather than acceleration. The phasic activation of whisking muscles is correlated with velocity14, so we speculate that self-motion responses after cutting can be explained by torque exerted internally by activation of specific muscles. Mouse whiskers are actuated by two groups of fast twitch muscles15: one intrinsic protractor (IP) and two extrinsic retractors (ER). Electromyogram (EMG) recordings in either or both of these muscle groups closely resembled phase tuning in recordings in which the whisker was fully cut (Figure 2C). The inertia hypothesis was partially correct. The first prediction was generally true because we observed acceleration tuning in a large portion of units (Figure 2D). Whisker cutting reduced spike rates in several units. Therefore, based on the results of these experiments and our mechanical framework, we propose that stresses related to inertial moment and activation of whisking muscles contribute to self-motion responses in primary afferents. Mechanical input to primary afferents likely underlies phase tuning observed in the whisker system. Phase coding may serve as an important proprioceptive signal during whisker exploration. References:

1. Johnson, K.O., and Hsiao, S.S. (1992). Neural mechanisms of tactual form and texture perception. Annu Rev Neurosci 15, 227-250.

2. Abraira, V.E., and Ginty, D.D. (2013). The sensory neurons of touch. Neuron 79, 618-639.

3. Bai, L., Lehnert, B.P., Liu, J., Neubarth, N.L., Dickendesher, T.L., Nwe, P.H., Cassidy, C., Woodbury, C.J., and Ginty, D.D. (2015). Genetic identification of an expansive mechanoreceptor sensitive to skin stroking. Cell 163, 1783-1795.

4. Pammer, L., O'Connor, D.H., Hires, S.A., Clack, N.G., Huber, D., Myers, E.W., and Svoboda, K. (2013). The mechanical variables underlying object localization along the axis of the whisker. J Neurosci 33, 6726-6741.

5. Campagner, D., Evans, M.H., Bale, M.R., Erskine, A., and Petersen, R.S. (2016). Prediction of primary somatosensory neuron activity during active tactile exploration. Elife 5.

6. Bush, N.E., Schroeder, C.L., Hobbs, J.A., Yang, A.E., Huet, L.A., Solla, S.A., and Hartmann, M.J. (2016). Decoupling kinematics and mechanics reveals coding properties of trigeminal ganglion neurons in the rat vibrissal system. Elife 5.

7. Fleming M.S., Luo W. (2013). The anatomy, function, and development of mammalian Aβ low-threshold mechanoreceptors. Front Biol (Beijing). 1;8(4).

8. Khatri, V., Bermejo, R., Brumberg, J.C., Keller, A., and Zeigler, H.P. (2009). Whisking in air: encoding of kinematics by trigeminal ganglion neurons in awake rats. J Neurophysiol 101, 1836-1846.

9. Moore, J.D., Mercer Lindsay, N., Deschenes, M., and Kleinfeld, D. (2015). Vibrissa self-motion and touch are reliably encoded along the same somatosensory pathway from brainstem through thalamus. PLoS Biol 13, e1002253.

10. Wallach, A., Bagdasarian, K., and Ahissar, E. (2016). On-going computation of whisking phase by mechanoreceptors. Nat Neurosci 19, 487-493.

11. Curtis, J.C., and Kleinfeld, D. (2009). Phase-to-rate transformations encode touch in cortical neurons of a scanning sensorimotor system. Nat Neurosci 12, 492-501.

12. Poulet J.F.A and Petersen C.C.H. (2008) Internal brain state regulates membrane potential synchrony in barrel cortex of behaving mice. Nature 454, 881-885.

13. Quist B.W., Seghete V., Huet L.A., Murphey T.D., Hartmann M.J. (2014) Modeling forces and moments at the base of a rat vibrissa during noncontact whisking and whisking against an object. J Neurosci 23;34(30):9828-44.

14. Hill, D.N., Bermejo, R., Zeigler, H.P., and Kleinfeld, D. (2008). Biomechanics of the vibrissa motor plant in rat: rhythmic whisking consists of triphasic neuromuscular activity. J Neurosci 28, 3438-3455.

15. Dorfl, J. (1985). The innervation of the mystacial region of the white mouse: A topographical study. J Anat 142, 173-184.

1

CURICULUM VITAE

Kaushal V. Asrani, M.B.B.S, Ph.D. Johns Hopkins, School of Medicine

1550 Orleans Street, CRB-2 Room 316, Baltimore, MD 21231. Phone: 410- 926-8061 (cell). Email: kasrani1@jhmi.edu

EDUCATION Ph.D. in Molecular Medicine (Cancer Biology) 2006-2012 University of Maryland School of Medicine, Baltimore, MD

M.B.B.S. (Bachelor of Medicine and Bachelor of Surgery) 1999-2005 Byramjee Jeejeebhoy Medical College and Sassoon General Hospital, Pune, India

PROFESSIONAL EXPERIENCE

Johns Hopkins, School of Medicine, Postdoctoral Fellow, Mentor: Dr. Tamara Lotan Aug 2012-present

Current projects:

x To determine the role of mTORC1 signaling in the regulation of epithelial barrier formation

x To determine the role of mTORC1 signaling in lysosomal biogenesis x To study the role of mTORC1 signaling in mammary development and branching

morphogenesis

University of Maryland School of Medicine, Center for Vascular and Inflammatory Diseases Graduate Research Assistant, Mentor: Dr. Jeffrey Winkles Oct 2007-July 2012

Studied the role of the Fn14 receptor in HER2/HER3-positive breast cancers, non-small cell lung cancer and melanoma

x Studied the regulation of Fn14 gene expression in HER2 and Heregulin-ȕ�-driven breast cancer cells and tumors using techniques such as Western blotting, qRT-PCR, FACS, luciferase-reporter assays

x Determined the contribution of Fn14 signaling to the invasive and proliferative phenotype of HER2 and Heregulin ȕ� overexpressing tumors using techniques such as MTT assays, clonogenic assays, mammosphere assays, Matrigel invasion assays, wound healing assays and tumor xenograft studies in mice

Sassoon General Hospital, Pune, India and Rural Training Center, Sirur, India Intern Jan 2004-Dec 2004

Obtained clinical experience while rotating in the departments of Internal Medicine, General Surgery, Pediatrics and Obstetrics and Gynaecology.

2

AWARDS AND HONORS “Best Oral Presentation Award” at the Second Annual Cancer Biology Research Retreat, University of Maryland School of Medicine, Baltimore, MD May 2011 Selected to attend the Sixth Annual NIH National Graduate Student Research Conference, at NIH, Bethesda, MD October 2011 Selected to attend the 2012 St. Jude National Graduate Student Symposium (NGSS) at the St. Jude Children's Research Hospital in Memphis, TN March 2012 Abstract was selected for mini-symposium presentation at the 2017 SID Annual Meeting in Portland, OR April 2017 Selected as one of the recipients of the 2017 SID Albert Kligman Travel Fellowship at the 2017 SID Annual Meeting in Portland, OR April 2017 Selected as one of the candidates for the first ever Regulatory Affairs Training Program conducted at the Johns Hopkins University Jan 2018 PROFESSIONAL MEMBERSHIPS American Association for Cancer Research 2009 – present American Society for Cell Biology 2009 – present Society for Investigative Dermatology 2017 – present

TECHNICAL SKILLS Biochemistry: Standard protein extraction, expression, isolation and detection, SDS-PAGE, Western blotting, co-immunoprecipitation, pulse-chase assays, surface biotinylation assays, cell fractionation, ELISA and luciferase reporter assays, Mass spectrometry and phospho proteomic analyses Molecular Biology: Quantitative real-time PCR (qRT-PCR), ChIP, RNAi techniques Cell Culture: Mammalian cell culture techniques, transient and stable transfections, FACS, MTT assays, clonogenic assays, mammary organoid cultures, murine keratinocyte cultures, Matrigel invasion assays, scratch wound-healing assays (migration assays), dispase dissociation assays, transmission electron microscopy, confocal microscopy, immunofluorescence, immunohistochemistry and fluorescence/DIC microscopy In-vivo Skills: Mice breeding, measuring and weighing tumors in mice, bioluminescence imaging in mice, skin grafting, maintenance of mouse colonies and genotyping. Presentation and Analysis: Proficient with Microsoft office, basic statistics and NCBI databases, Analyses of Microarray datasets

3

OTHER SKILLS Ability to collaborate: Collaborated with UMB faculty at the Greenebaum Cancer Center (Dr. Rena Lapidus for mouse work), the CVID (FACS, microscopy) and TGen, a research institute in Phoenix, Arizona for technical knowledge, skills and equipment use. Collaborated with Johns Hopkins faculty Dr. Pierre Coulombe (mouse keratinocyte studies) and Dr. Akhilesh Pandey (Mass spectrometry and phospoproteomic analyses). Team leadership: Effectively managed my own projects as well as delegated work for research technicians and students; served on Organizing Committee for the Second Annual Cancer Biology Research Retreat at the University of Maryland. Teaching: Trained and supervised numerous graduate students and summer interns both at a technical level and also day-to-day activities to meet project deadlines. This has lead to successful completion of their projects. Analytical and communication skills: Analyzed complex data sets critically, and presented findings at departmental seminars, program retreats and poster presentations at conferences. This has also lead to paper publications, preliminary data for grants and grant progress reports. Grant writing skills: Applied to a Department of Defense (DOD) FY10 Breast Cancer Research Program (BCRP) for the Predoctoral Traineeship Award and received a Global score of 1.8 (Excellent) in the First Tier-Scientific Peer Review. PEER-REVIEWED PUBLICATIONS D'Souza DR, Salib MM, Bennett J, Mochin-Peters M, Asrani K, Goldblum SE, Renoud KJ, Shapiro P, and Passaniti A. (2009). Hyperglycemia regulates RUNX2 activation and cellular wound healing through the aldose reductase polyol pathway. J Biol Chem 284: 17947-55. Whitsett TG, Cheng E, Inge L, Asrani K, Jameson N, Hostetter G, Weiss GJ, Kingsley C, Loftus J, Bremner R, Tran NL, Winkles JA. (2012). Elevated expression of Fn14 in non-small cell lung cancer correlates with activated EGFR and promotes tumor cell migration and invasion. Am J Pathol 181: 111-20.

Zhou H, Ekmekcioglu S, Marks JW, Mohamedali KA, Asrani K, Phillips KK, Brown SAN, Cheng E, Weiss MB, Hittelman WN, Tran NL, Yagita H, Winkles JA and Rosenblum MG. (2012). The TWEAK receptor Fn14 is a therapeutic target in melanoma: immunotoxins targeting Fn14 receptor for malignant melanoma treatment. J Invest Dermatol 133: 1052-62.

Asrani K, Keri RA, Galisteo R, Brown SA, Morgan SJ, Ghosh A, Tran NL, Winkles JA. (2013). The HER2- DQG�+HUHJXOLQ�ȕ�� �+5*�-Inducible TNFR Superfamily Member Fn14 Promotes HRG-Driven Breast Cancer Cell Migration, Invasion, and MMP9 Expression. Mol Cancer Res 11: 393-404.

Asrani K, Sood A, Torres A, Georgess D, Phatak P, Kaur H, Dublin A, Talbot CC Jr, Elhelu L, Ewald AJ, Xiao B, Worley P, Lotan TL. mTORC1 loss impairs epidermal adhesion via TGF-ȕ�5KR�NLQDVH�activation. J Clin Invest 127: 4001-4017.

Asrani K, Sood A, Lotan TL. mTORC1-induced Akt inhibition enhances TFEB-family driven lysosomal biogenesis. (Manuscript in preparation).

4

Torres A, Tsai H, Rooper L, Asrani K, Westra W, Lotan TL. Gene Expression Meta-Signature Identifies INSM1 Expression and YAP1 Loss as Markers of Neuroendocrine Prostate Cancer. (Manuscript in preparation).

Noë M, Rezaee N, Asrani K, Skaro M, Groot VP, Wu PH, Olson M, Hong SM, Kim SJ, Weiss MJ, Wolfgang CL, Makary MA, He J, Cameron JL, Wirtz D, Offerhaus GJA, Brosens LAA, Wood LD, Hruban RH. Immunolabeling of Cleared Human Pancreata Provides Insights into the Three-dimensional Pancreatic Anatomy and Pathology. (Manuscript submitted to the Journal of Pathology).

POSTER PRESENTATIONS D'Souza DR, Salib MM, Bennett J, Mochin-Peters M, Asrani K, Goldblum SE, Renoud KJ, Shapiro P, and Passaniti A. (2009). Hyperglycemia regulates RUNX2 activation and cellular wound healing through the aldose reductase polyol pathway.

100th Annual Meeting of the AACR, Denver, CO, April 18-23, 2009.

Asrani KV, Winkles JA. (2010). HER2 signaling regulates Fn14 receptor expression in breast cancer cells.

101st Annual Meeting of the AACR, Washington, D.C., April 17-21, 2010. 50th ASCB Annual Meeting, Philadelphia, PA, December 11-15, 2010. Tran NL, Asrani K, Jameson N, Inge L, Cheng L, Hostetter G, Weiss GJ, Tuncali S, Bremner R, Winkles JA. (2011). The Fn14 receptor is highly expressed in NSCLC tumors and in NSCLC cell lines harboring EGFR-activating mutations.

102nd Annual Meeting of the AACR, Orlando, FL, April 2-6, 2011.

Zhou H, Marks JW, Yagita J, Asrani K, Winkles JA and Rosenblum MG. (2011). Development and characterization of Fn14 receptor-targeted immunotoxins for breast cancer therapy.

DOD BCRP Era of Hope Conference, Orlando, FL, August 2-5, 2011.

Zhou H, Ekmekcioglu S, Marks JW, Asrani K, Phillips KK, Brown SAN, Cheng E, Weiss MB, Hittelman WN, Tran NL, Yagita H, Winkles JA, Rosenblum MG. (2012). TWEAK Receptor (Fn14) is a novel target in Melanoma: Characterization of unique targeted therapeutics.

103rd Annual Meeting of the AACR, Chicago, IL, March 31-April 4, 2012.

5

Asrani K, Sood A, Torres A, Dublin A, Talbot CC Jr, Elhelu L, Xiao B, Worley P, Lotan TL. (2016). mTOR signaling regulates epidermal barrier function.

22nd Annual Fellow Research Day, Johns Hopkins Kimmel Cancer Center (SKCCC), Baltimore, MD, May 24, 2016.

Asrani K, Sood A, Torres A, Dublin A, Talbot CC Jr, Elhelu L, Xiao B, Worley P, Lotan TL. (2017). mTORC1 loss impairs epidermal cell-cell adhesion and differentiation via rho kinase activation downstream of TGF-ȕ�VLJQDOLQJ�

Department of Dermatology Research Day, Johns Hopkins Kimmel Cancer Center (SKCCC), Baltimore, MD, May 24, 2017.

Asrani K, Sood A, Torres A, Dublin A, Talbot CC Jr, Elhelu L, Xiao B, Worley P, Lotan TL. (2017). mTORC1 loss impairs epidermal cell-cell adhesion and differentiation via rho kinase activation downstream of TGF-ȕ�VLJQDOLQJ�

76th Annual Meeting of the Society of Investigative Dermatology, Portland, OR, April 26-29, 2017.

mTORC1 Loss Impairs Epidermal Adhesion via TGF-ȕ�5KR�.LQDVH�$FWLYDWLRQ�

Abstract: The mammalian target of rapamycin complex 1 (mTORC1) is a master regulator of normal homeostatic, as well as pathological oncogenic signaling; however, its role in epithelial development has not been studied extensively in vivo. Here, we use neonatal mouse epidermis as a model system to study the cellular effects and signaling feedback sequelae of mTORC1 loss-of-function in epithelial tissue. In mice with epidermal-specific deletion of mTORC1 components Rheb or Rptor, we find that mTORC1 loss-of-function unexpectedly results in a profound skin barrier defect with epidermal abrasions, suprabasal blistering and early postnatal lethality, due to a thinned epidermis with decreased desmosomal protein and RNA expression impaired stratification and incomplete biochemical differentiation. Mechanistically, impaired desmosomal cellular adhesion and biochemical differentiation in the setting of mTORC1 loss-of-function is caused by constitutive activation of Rho kinase (ROCK) signaling with a resulting increase in actin cytoskeletal tension. Furthermore, inhibition or silencing of ROCK1 is sufficient to rescue keratinocyte adhesion and biochemical differentiation in vitro. mTORC1 loss-of-function also results in marked feedback up-regulation of upstream TGF-ȕ� VLJQDOLQJ��thereby triggering an increase in ROCK activity and its downstream effects on desmosomal gene expression. These findings highlight a previously undescribed role for mTORC1 in the regulation of epidermal barrier function, actin cytoskeletal dynamics, desmosomal gene regulation and cell adhesion and underscore the complexity of signaling feedback after mTORC1 inhibition.

Introduction: The mTOR signaling pathway plays a critical role in integrating nutrient and growth

factor inputs to regulate cell proliferation, growth and metabolism. mTORC1 activation requires the small Ras-related GTPase, Rheb (1), as well as an adaptor protein, Raptor (2). Despite its well-characterized role in cellular processes that facilitate tumorigenesis, mTORC1 inhibitors have been ineffective for the treatment of most solid cancers (3), potentially due to relief of mTORC1-induced feedback inhibition of upstream growth factor receptor signaling (4-9). Furthermore, the in vivo effects of mTORC1 inhibition have not been well-studied in epithelia and tumors and need to be fully characterized before we can devise strategies to circumvent tumor resistance to mTORC1 inhibition. Among epithelial tissues, the skin is an excellent model system to study mTOR function since mTORC1 inhibitor treatment in transplant or cancer patients is associated with impaired wound healing and mild to life-threatening skin toxicity (10-12). In the skin, intercellular adhesion--via adherens junctions, desmosomes and tight junctions--as well as terminal differentiation of keratinocytes, are essential for epidermal barrier formation (13). Desmosomes are calcium-dependent adhesion molecules that anchor to tension-bearing keratin intermediate filaments and provide a rigid framework for the cell. As a result, loss of desmosomal components in humans and mice by genetic or autoimmune mechanisms leads to a spectrum of epidermal barrier defects (14-18).

The actin cytoskeleton is central to cell-cell adhesion. Adherens junction assembly requires a fine balance in the activities of two members of the Rho family of small GTPases, Rac and Rho, and of the downstream effectors of Rho, mDia and Rho kinase (ROCK) (19-22). In turn, ROCK signaling and the resulting increase in actomyosin contractility likely regulate desmosomal junction formation at multiple levels (23). While appropriate Rho localization and actomyosin contractility are required for satisfactory incorporation of desmosomes, sustained Rho-ROCK activity can hamper this process in vitro (24). In addition, cytoskeletal contractility may exert complex effects on downstream gene expression through mechanotransduction, affecting the activity of multiple transcription factors (25). Despite this wealth of in vitro data, relatively little is known about the in vivo requirement of actomyosin contractility for desmosomal adhesion and epidermal barrier formation.

mTOR signaling perturbation in the epidermis has been associated with skin carcinogenesis (26), stem cell senescence (27, 28) and delayed wound healing (29). Here, we used two separate genetic models to examine the consequences of conditional epidermal mTORC1 loss-of-function. Unexpectedly, we find that mTORC1 signaling in the epidermis is required for desmosomal cell-cell adhesion and keratinocyte biochemical differentiation, as exemplified by epidermal blistering, defective skin barrier function and neonatal lethality of the cKO mice. Mechanistically, these findings are a direct consequence of heightened cytoskeletal tension due to increased ROCK activity, resulting in impaired desmosomal gene expression and adherens junction maturation. Probing further upstream, mTORC1 loss leads to marked feedback up-regulation of TGF-ȕ� VLJQDOLQJ� DQG� GRZQVWUHDP� HIIHFWV� RQ� F\WRVNHOHWDO� FRQWUDFWLOLW\�� FHOO�adhesion and biochemical differentiation. These studies reveal a novel in vivo role of mTORC1 function in epithelial biology that has relevance for human disease. Figure Legends:

Figure 1: Epidermal-specific mTORC1 loss-of-function models have skin barrier defects and evidence of impaired cell-cell adhesion. (A) Immunoblotting of WT and Rheb cKO P0 (day 0) epidermis for Rheb expression. (B) Immunofluorescence of WT and Rheb cKO E18.5

epidermis for p-S6 as a marker of mTORC1 activity. (C) Rheb cKO pups are smaller at P0, with thin, shiny skin (top left) and fail to exclude toluidine blue dye at E18.5 (top right), consistent with an epidermal barrier defect. Grossly visible abrasions (arrow; bottom right) and a prominent vascular pattern are seen in P0 cKO epidermis. (D) Representative histologic backskin sections from Rheb cKO P0 pups reveal markedly thinned epidermis with loss of the upper keratinizing cell layers (right). (E) Immunoblotting of WT and Rptor cKO P0 epidermis for Raptor expression. (F) Rptor cKO pups have visible epidermal blisters at P0 (right, arrow). (G) Rptor cKO pups have a thinned epidermis by histology (right half of panel, H&E staining) with visible suprabasal blistering (denoted by *). (H) Transmission electron microscopy (TEM) of Rheb cKO P0 epidermis (first row, E) demonstrates diminished granular layer (G) and intercellular gaps (first row, arrows). Desmosomes (second row, arrowheads) are evident (second row, arrow). Tonofilaments are collapsed (second and third rows, arrows) and not easily visible (third row, arrowheads) in cKO cells. Scale bar = 4 µm (top row) and 2 µm (middle/bottom) row. (I) Desmosomes in Rptor cKO P0 epidermis have diminished keratin attachment by TEM (left; Scale bar = 500 nm). (J) Desmosomes in Rheb cKO keratinocytes have less dense cytoplasmic plaques (arrows) and poorly formed midline (arrowhead) (left panels; scale bar = 100 nm). (K) Dispase dissociation for epithelial fragments (arrows) in Rheb WT and cKO keratinocyte monolayers (left panel). Scale bar = 1 cm. (L) Immunoblotting for junctional proteins in Rptor WT vs. cKO P0 epidermis. (M) Immunofluorescence for desmosomes in Rheb WT and cKO keratinocytes (top panels), with intracellular gaps (arrow). Scale bar=50 µm (top two panels) and 10 µm (bottom panel). 3rd panel of merged staining is inset from boxed area in top 2 panels.

Figure 2: Impaired desmosomal cell adhesion with epidermal mTORC1 loss-of-function is associated with increased cytoskeletal tension and Rho kinase (ROCK) activity and is rescued by ROCK inhibition or depletion. (A) Immunofluorescence in Rheb and Rptor WT and cKO keratinocytes for desmosomal, focal adhesion (vinculin) and adherens junction markers (E-cadherin, phalloidin staining for actin filaments). 2nd panel of vinculin staining is inset from boxed area in 1st panel. Scale bar = 50 µm (1st panel), 25 µm (2nd panel) 10 µm (3rd panel) and 50 µm (4th panel). (B) Immunoblotting of Rptor WT and cKO P0 epidermal lysates for ROCK activation. (C) Immunoblotting of Rheb WT and cKO keratinocyte lysates for ROCK activation. (D) Immunoblotting of WT and inducible Rptor cKO cell lysates with or without ROCK inhibition using Y27632 (2, 20 µM) or fasudil (5, 50 µM) for desmosomal components and ROCK activity markers. (E) Immunofluorescence for adherens junction components (top panels; left) and DSP1/2 (middle panels; left) in Rheb cKO keratinocytes following Y27632 (10 µM) treatment. TEM (bottom panels; left) of Rheb cKO desmosomes with or without Y27632 treatment. Scale bar = 50 µm (top two panels; left) and 500 nm (bottom panels; left). Quantification of DSP1/2 immunofluorescence (middle panel; right) and E-cadherin (top panel; right) (r=3; n> 33 cell borders each) in Rheb cKO keratinocytes ± Y27632 (p-value by Mann-Whitney test (DSP) or Student’s T-test (E-cadherin)). Quantification of desmosomal length on TEM for Rheb cKO keratinocytes ± Y27632 (bottom panel; right) (bar represents mean length in nm; n=11 and 12; p-values by Student’s T-test). (F) Dispase dissociation (left panel) for Rheb WT or cKO keratinocytes ± Y27632 (20 µM). Scale bar = 1 cm. Quantification of dispase assay experiments (right panel) (r=4; p-values by one-way ANOVA). (G) Immunoblotting of Rheb WT or cKO keratinocytes ± Y27632 (20 µM) for biochemical differentiation markers. (H) Immunoblotting of Rheb cKO cells ± ROCK1 siRNA (50nm) for ROCK1 and markers of ROCK activity, desmosomal and biochemical differentiation markers. Error bars represent SEM throughout.

A

F

C

E

BDG

H

J I

KL

M

Figure 1.

Figure 2.A

BC

DE

FG

H

References:

1. Inoki K, Li Y, Xu T, and Guan KL. Rheb GTPase is a direct target of TSC2 GAP activity and regulates mTOR signaling. Genes Dev. 2003;17(15):1829-34.

2. Kim DH, Sarbassov DD, Ali SM, King JE, Latek RR, Erdjument-Bromage H, Tempst P, and Sabatini DM. mTOR interacts with raptor to form a nutrient-sensitive complex that signals to the cell growth machinery. Cell. 2002;110(2):163-75.

3. Guertin DA, and Sabatini DM. The pharmacology of mTOR inhibition. Sci Signal. 2009;2(67):pe24.

4. Shah OJ, Wang Z, and Hunter T. Inappropriate activation of the TSC/Rheb/mTOR/S6K cassette induces IRS1/2 depletion, insulin resistance, and cell survival deficiencies. Curr Biol. 2004;14(18):1650-6.

5. Harrington LS, Findlay GM, Gray A, Tolkacheva T, Wigfield S, Rebholz H, Barnett J, Leslie NR, Cheng S, Shepherd PR, et al. The TSC1-2 tumor suppressor controls insulin-PI3K signaling via regulation of IRS proteins. J Cell Biol. 2004;166(2):213-23.

6. Zhang H, Bajraszewski N, Wu E, Wang H, Moseman AP, Dabora SL, Griffin JD, and Kwiatkowski DJ. PDGFRs are critical for PI3K/Akt activation and negatively regulated by mTOR. J Clin Invest. 2007;117(3):730-8.

7. Carracedo A, Ma L, Teruya-Feldstein J, Rojo F, Salmena L, Alimonti A, Egia A, Sasaki AT, Thomas G, Kozma SC, et al. Inhibition of mTORC1 leads to MAPK pathway activation through a PI3K-dependent feedback loop in human cancer. J Clin Invest. 2008;118(9):3065-74.

8. Hsu PP, Kang SA, Rameseder J, Zhang Y, Ottina KA, Lim D, Peterson TR, Choi Y, Gray NS, Yaffe MB, et al. The mTOR-regulated phosphoproteome reveals a mechanism of mTORC1-mediated inhibition of growth factor signaling. Science. 2011;332(6035):1317-22.

9. Yu Y, Yoon SO, Poulogiannis G, Yang Q, Ma XM, Villen J, Kubica N, Hoffman GR, Cantley LC, Gygi SP, et al. Phosphoproteomic analysis identifies Grb10 as an mTORC1 substrate that negatively regulates insulin signaling. Science. 2011;332(6035):1322-6.

10. Valente JF, Hricik D, Weigel K, Seaman D, Knauss T, Siegel CT, Bodziak K, and Schulak JA. Comparison of sirolimus vs. mycophenolate mofetil on surgical complications and wound healing in adult kidney transplantation. Am J Transplant. 2003;3(9):1128-34.

11. Liu LS, McNiff JM, and Colegio OR. Palmoplantar peeling secondary to sirolimus therapy. Am J Transplant. 2014;14(1):221-5.

12. Balagula Y, Rosen A, Tan BH, Busam KJ, Pulitzer MP, Motzer RJ, Feldman DR, Konner JA, Reidy-Lagunes D, Myskowski PL, et al. Clinical and histopathologic characteristics of rash in cancer patients treated with mammalian target of rapamycin inhibitors. Cancer. 2012;118(20):5078-83.

13. Candi E, Schmidt R, and Melino G. The cornified envelope: a model of cell death in the skin. Nature reviews Molecular cell biology. 2005;6(4):328-40.

14. Delva E, Tucker DK, and Kowalczyk AP. The desmosome. Cold Spring Harb Perspect Biol. 2009;1(2):a002543.

15. Chidgey M, Brakebusch C, Gustafsson E, Cruchley A, Hail C, Kirk S, Merritt A, North A, Tselepis C, Hewitt J, et al. Mice lacking desmocollin 1 show epidermal fragility

accompanied by barrier defects and abnormal differentiation. J Cell Biol. 2001;155(5):821-32.

16. Vasioukhin V, Bowers E, Bauer C, Degenstein L, and Fuchs E. Desmoplakin is essential in epidermal sheet formation. Nat Cell Biol. 2001;3(12):1076-85.

17. Sumigray KD, and Lechler T. Cell adhesion in epidermal development and barrier formation. Current topics in developmental biology. 2015;112(383-414.

18. Koch PJ, Mahoney MG, Ishikawa H, Pulkkinen L, Uitto J, Shultz L, Murphy GF, Whitaker-Menezes D, and Stanley JR. Targeted disruption of the pemphigus vulgaris antigen (desmoglein 3) gene in mice causes loss of keratinocyte cell adhesion with a phenotype similar to pemphigus vulgaris. J Cell Biol. 1997;137(5):1091-102.

19. Takeichi M. Dynamic contacts: rearranging adherens junctions to drive epithelial remodelling. Nature reviews Molecular cell biology. 2014;15(6):397-410.

20. Zandy NL, Playford M, and Pendergast AM. Abl tyrosine kinases regulate cell-cell adhesion through Rho GTPases. Proc Natl Acad Sci U S A. 2007;104(45):17686-91.

21. Sahai E, and Marshall CJ. ROCK and Dia have opposing effects on adherens junctions downstream of Rho. Nat Cell Biol. 2002;4(6):408-15.

22. Fang WB, Ireton RC, Zhuang G, Takahashi T, Reynolds A, and Chen J. Overexpression of EPHA2 receptor destabilizes adherens junctions via a RhoA-dependent mechanism. J Cell Sci. 2008;121(Pt 3):358-68.

23. Godsel LM, Hsieh SN, Amargo EV, Bass AE, Pascoe-McGillicuddy LT, Huen AC, Thorne ME, Gaudry CA, Park JK, Myung K, et al. Desmoplakin assembly dynamics in four dimensions: multiple phases differentially regulated by intermediate filaments and actin. J Cell Biol. 2005;171(6):1045-59.

24. Godsel LM, Dubash AD, Bass-Zubek AE, Amargo EV, Klessner JL, Hobbs RP, Chen X, and Green KJ. Plakophilin 2 couples actomyosin remodeling to desmosomal plaque assembly via RhoA. Mol Biol Cell. 2010;21(16):2844-59.

25. Rajakyla EK, and Vartiainen MK. Rho, nuclear actin, and actin-binding proteins in the regulation of transcription and gene expression. Small GTPases. 2014;5(e27539.

26. Lu ZH, Shvartsman MB, Lee AY, Shao JM, Murray MM, Kladney RD, Fan D, Krajewski S, Chiang GG, Mills GB, et al. Mammalian target of rapamycin activator RHEB is frequently overexpressed in human carcinomas and is critical and sufficient for skin epithelial carcinogenesis. Cancer Res. 2010;70(8):3287-98.

27. Castilho RM, Squarize CH, Chodosh LA, Williams BO, and Gutkind JS. mTOR mediates Wnt-induced epidermal stem cell exhaustion and aging. Cell Stem Cell. 2009;5(3):279-89.

28. Iglesias-Bartolome R, Patel V, Cotrim A, Leelahavanichkul K, Molinolo AA, Mitchell JB, and Gutkind JS. mTOR inhibition prevents epithelial stem cell senescence and protects from radiation-induced mucositis. Cell Stem Cell. 2012;11(3):401-14.

29. Feldmeyer L, Hofbauer GF, Boni T, French LE, and Hafner J. Mammalian target of rapamycin (mTOR) inhibitors slow skin carcinogenesis, but impair wound healing. Br J Dermatol. 2012;166(2):422-4.

The Reck-Gpr124 complex activates Wnt signaling to promote CNS angiogenesis Chris Cho1, Philip M. Smallwood1,2, Jeremy Nathans1,2,3,4

1Department of Molecular Biology and Genetics, Johns Hopkins University School of Medicine, Baltimore, Maryland, 21205 2Howard Hughes Medical Institute, Johns Hopkins University School of Medicine, Baltimore, MD, 21205 3Department of Neuroscience, Johns Hopkins University School of Medicine, Baltimore, Maryland, 21205 4Department of Ophthalmology, Johns Hopkins University School of Medicine, Baltimore, Maryland, 21205

Central nervous system (CNS) angiogenesis and blood-brain barrier (BBB) development and maintenance are controlled by canonical Wnt signaling. Specifically, Wnt7a and Wnt7b are required for normal vascular development in the forebrain and spinal cord. Two cell-surface proteins – orphan receptor Gpr124, and more recently, GPI-anchored Reck (reversion-inducing cysteine-rich protein with kazal motifs) – have emerged as specific co-activators of the Wnt7a/7b signaling axis. Whether this novel function of Reck plays a role in mammalian vascular development and how these two co-activators might cooperate to drive angiogenesis remain unclear. Here we report that Reck promotes CNS angiogenesis by activating the canonical Wnt pathway in mice. We identify the critical domains of both Reck and Gpr124 that are required for Wnt activity, and demonstrate that these regions are important for direct binding and complex formation. Importantly, weakening this interaction by targeted mutagenesis reduces Reck-Gpr124 stimulation of Wnt7a signaling in cell culture and impairs CNS angiogenesis. Finally, a soluble Gpr124 probe binds specifically to cells expressing Frizzled (Fz), Wnt7a or Wnt7b, and Reck; and a soluble Reck probe binds specifically to cells expressing Fz, Wnt7a or Wnt7b, and Gpr124. Altogether, these experiments reveal an essential role for Reck and Gpr124 as integral parts of the cell-surface protein complex that transduces Wnt7a- and Wnt7b-specific signals in mammalian endothelial cells to promote CNS angiogenesis and BBB formation and maintenance.

Chris Cho

The Reck-Gpr124 complex activates Wnt signaling to promote CNS angiogenesis

Canonical Wnt signaling plays a pivotal role in promoting central nervous system (CNS) angiogenesis and blood-brain barrier (BBB) formation and maintenance. Specifically, Wnt7a and Wnt7b are required for vascular development in the forebrain and ventral spinal cord[1,2]. Yet, how these two ligands – among the 19 mammalian Wnts – are selectively communicated to Frizzled receptors expressed on endothelial cells remains largely unclear. Here, I report a novel paradigm for Wnt specificity. I have identified two endothelial cell surface proteins – GPI-anchored Reck and orphan receptor Gpr124 – as essential receptor co-factors that assemble into a multi-protein complex with Wnt7a/7b and Frizzled for the development of the mammalian neurovasculature.

In previous work, we (and others) have shown that Gpr124 controls CNS angiogenesis in the

developing mouse by activating the Wnt7a/7b signaling pathway[3-5]. In an attempt to identify novel genes that may be cooperating with Gpr124, our collaborators Vanhollebeke et al. conducted a morpholino knockdown screen in zebrafish[5]. Remarkably, knockdown of one candidate gene – reversion-inducing cysteine-rich protein with Kazal motifs (Reck) – phenocopied the Gpr124 null zebrafish (i.e. cerebral vascular defects)[5]. Given the phenotypic similarities, I tested whether Reck, like Gpr124, could stimulate the Wnt7a/7b pathway. Using a reporter cell line for canonical Wnt signaling (Super Top Flash; STF), I showed that Reck and Gpr124 synergistically activate the canonical pathway in the presence of Wnt7a or Wnt7b, but not for any of the other 19 Wnts, including Norrin, an atypical non-Wnt ligand (Figure 1a)[5]. Reck has been extensively characterized as a matrix metalloproteinase (MMP) inhibitor; however, a connection to canonical Wnt signaling had not been described[6-8]. To determine whether this novel Wnt function of Reck plays a role in mammalian neurovascular development, I generated both knockout (ReckΔex2/Δex2) and endothelial cell-specific knockout (data not shown) mice. At embryonic day 13.5 (E13.5), both mutant animals exhibit hemorrhaging in the forebrain and spinal cord, and display significantly reduced vascular density accompanied by increased infiltration of macrophages in these CNS regions, phenotypes that closely match the Gpr124 KO embryos (Figures 1B and 1C). In addition to these defects, I also observed a compromised BBB in Reck mutant mice (data not shown). If Reck is in fact responsible for activating canonical Wnt signaling to drive these developmental processes, then artificially stabilizing the downstream signal transducer, β-catenin, in endothelial cells should rescue these phenotypes. Since exon 3 of the β-catenin gene (Ctnnb1) harbors glycogen synthase kinase-3 (GSK3) phosphorylation sites, which target β-catenin for its degradation, I used a conditional β-catenin allele (Ctnnb1flex3) to delete this region. In this genetic rescue experiment, I showed that Reck knockout animals with stabilized β-catenin in endothelial cells showed minimal to no hemorrhaging in the forebrain and spinal cord and a near-complete restoration of the neurovasculature (Figure 1D). I therefore conclude that Reck plays a critical role in activating canonical Wnt signaling to control CNS angiogenesis and regulate the BBB. In a comprehensive series of mouse genetic experiments, I was also able to show that Reck genetically interacts with Gpr124 and Wnt7a/Wnt7b in promoting CNS angiogenesis (data not shown). The strong genetic evidence, in addition to the specific and synergistic in vitro Wnt7a/7b activation (Figure 1a), points to a cooperative role between Reck and Gpr124. To refine their molecular mechanism, I tested domain mutants of both proteins using our in vitro Wnt reporter assay, and identified the N-terminal domains of both proteins to be of critical importance: Reck CC domains 1-5 (CC1-5) and Gpr124 Leucine Rich Repeat (LRR) and Immunoglobulin-like (Ig) domains (data not shown). [Note: The Reck CC domains share no homology to other domains, and were named ‘CC’ because of the presence of 5x dicysteine motifs].

Chris Cho

Having identified the critical domains required for Reck and Gpr124 to activate Wnt7a/7b signaling, I generated recombinant proteins of these domains fused to alkaline phosphatase (AP). These AP probes also contained a cartilage oligomeric matrix protein (COMP) domain to allow for pentamerization and increased avidity of the probes. The Reck(CC1-5) COMP-AP probe bound robustly to cells transfected with Gpr124, but not to its close homolog, Gpr125, which does not activate Wnt7a/7b signaling (Figure 1E). Additionally, deletion of the LRR, Ig, LRR+Ig, and HormR domains of Gpr124 completely abolished the ability of the Reck probe to bind (Figure 1E). To test direct binding between Reck and Gpr124 in a cell-free system, I generated Fc baits of Gpr124(LRR-Ig) and Reck(CC1-5), which were both fused to the constant region of human IgG. Upon capturing baits on Protein-G-coated wells, I showed that the Reck(CC1-5) COMP-AP probe bound specifically to the Gpr124 Fc bait (Figure 1F). Furthermore, the Reck(CC1-2) and Reck(CC1) COMP-AP probes were sufficient to promote binding (Figure 1F). To identify the key residues within Reck CC1 that were facilitating this interaction with Gpr124, I performed alanine scanning mutagenesis in this region (Figure 1G). Of the 16 COMP-AP mutant probes, seven showed readily detectable expression levels: Ala Scan-2, -7, -9, -10, -14, -15, and -16. All mutant probes bound to the Gpr124 Fc bait, except for Reck Ala Scan-9 and -10 (Figure 1H). Single alanine substitution of the five residues encompassed by these two adjacent block substitutions showed that three of the five – R69A, P71A, and Y73A – greatly reduced binding (Figure 1H). These three residues are highly conserved across vertebrates, whereas adjacent residues that are dispensable for binding are not. Importantly, full-length Reck derivatives with alanine substitutions at this binding site resulted in reduced Wnt7a/7b activation in vitro (Figure 1I). Furthermore, I used CRISPR-Cas9 to generate mice carrying a Reck allele with these alanine substitutions, and showed that disrupting this interface in vivo results in impaired CNS angiogenesis (Figure 1J). Altogether, these data support a model in which Reck-Gpr124 binding is important for Wnt7a/Wnt7b signaling during CNS angiogenesis.

What remains unclear is how Reck and Gpr124 operate with Wnt7a/7b and Frizzled. To

address this question, I generated Reck(CC1-5)-AP and AP-Gpr124(LRR-Ig) probes, lacking the COMP pentamerization domain. In this case, the Reck probe could not bind to cells transfected with Gpr124 (and vice versa), as their interaction is below the detection limit of the assay (Figure 2A). However, robust binding of the probes could be observed with the additional co-transfection of Wnt7a, Frizzled4 (Fz4), and Lrp5 (Wnt co-receptor) (Figure 2A). This complex formation was specific to Wnt7a and Wnt7b, as the probes could not bind in the presence of any of the other 17 Wnt ligands or Norrin (Figure 2B). These binding specificities precisely mirror the signaling specificity of Reck and Gpr124 previously defined in our in vitro reporter assay (Figure 1A). To assess Frizzled specificity, I showed that Reck and Gpr124 probes could bind to seven of ten Frizzleds, with only Fz3, Fz6, and Fz10 showing no detectable binding (Figure 2C). It is interesting to note that Fz3 and Fz6 do not activate canonical Wnt signaling, and are proposed to play a non-canonical role in planar cell polarity (PCP).

In the 20 years since the discovery that Frizzleds constitute the principal receptors for Wnts,

and the contemporaneous appreciation that the mammalian genome encodes for 10 Frizzleds and 19 Wnt ligands, the extent of Wnt-Frizzled specificity and the biological roles that such specificity might play have largely remained open questions. Here, I present a novel model for Wnt specificity with important biological function: Reck and Gpr124 are integral parts of the cell surface protein complex that transduces Wnt7a- and Wnt7b-specific signals in mammalian CNS endothelial cells to promote angiogenesis and regulate the BBB (Figure 2D). These findings signify the exciting possibility that receptor co-factors like Reck and Gpr124 exist for the other 17 Wnt ligands, and identifying these factors will be an important future direction for the field.

Chris Cho

References 1. Stenman, J.M., Rajagopal, J., Carroll, T.J., Ishibashi, M., McMahon, J., and McMahon, A.P. (2008).

Canonical Wnt Signaling Regulates Organ-Specific Assembly and Differentiation of CNS Vasculature. Science 322, 1247-1250.

2. Daneman, R., Agalliu, D., Zhou, L., Kuhnert, F., Kuo, C.J., and Barres, B.A. (2009). Wnt/beta-catenin

signaling is required for CNS, but not non-CNS, angiogenesis. Proc. Natl. Acad. Sci. USA 106, 6422-6422.

3. Zhou, Y., and Nathans, J. (2014). Gpr124 controls CNS angiogenesis and blood-brain barrier integrity by

promoting ligand-specific canonical Wnt signaling. Dev. Cell 31, 248-256.

4. Posokhova, E., Shukla, A., Seaman, S., Volate, S., Hilton, M.B., Wu, B., Morris, H., Swing, D.A., Zhou, M., Zudaire, E., et al. (2015). GPR124 functions as a WNT7-specific coactivator of canonical β-catenin signaling. Cell Rep. 10, 123-130.

5. Vanhollebeke, B., Stone, O.A., Bostaille, N., Cho, C., Zhou, Y., Maquet, E., Gauquier, A., Cabochette,

P., Fukuhara, S., Mochizuki, N., et al. (2015). Tip cell-specific requirement for an atypical Gpr124- and Reck-dependent Wnt/β-catenin pathway during brain angiogenesis. Elife 4, e06489.

6. Takahashi, C., Sheng, Z., Horan, T.P., Kitayama, H., Maki, M., Hitomi, K., Kitaura, Y., Takai, S.,

Sasahara, R.M., Horimoto, A., et al. (1998). Regulation of matrix metalloproteinase-9 and inhibition of tumor invasion by the membrane-anchored glycoprotein RECK. Proc. Natl. Acad. Sci. USA 95, 13221-13226.

7. Oh, J., Takahashi, R., Kondo, S., Mizoguchi, A., Adachi, E., Sasahara, R.M., Nishimura, S., Imamura,

Y., Kitayama, H., Alexander, D.B., et al. (2001). The membrane-anchored MMP inhibitor RECK is a key regulator of extracellular matrix integrity and angiogenesis. Cell 107, 789-800.

8. Chang, C.K., Hung, W.C., and Chang, H.C. (2008). The Kazal motifs of RECK protein inhibit MMP-9

secretion and activity and reduce metastasis of lung cancer cells in vitro and in vivo. J. Cell Mol. Med. 12, 2781-2789.

b

15

0

5

10

Re

lativ

e lu

cife

rase

act

ivity

30

10

20

00

200

400

800

600

Reck + Gpr124

Reck

Gpr124

pRK5 vector

Wn

t1

Wn

t2

Wn

t3

Wn

t2B

Wn

t3A

Wn

t4

Wn

t5A

Wn

t5B

Wn

t6

Wn

t7A

Wn

t7B

Wn

t8A

Wn

t8B

Wn

t9A

Wn

t9B

Wn

t10

A

Wn

t10

B

Wn

t11

Wn

t16

No

rrin

No

lig

an

d

AW

T

B

Bright-field

E13.5

E

EVTGGLAPGSAGAVCCNHSKDNQMCRDVCEQIFSSKSESRLKHLLQRAPDYCPETMVEIWSCMNSSLPGVFKKSDGWVGLG

H

mouse Reck CC130. . . . . . . .

A A A

1

AAAA

2AAAA3

AAA

4

AAAA

5AAAA6

AAAA

7AAAA8

AA A

9AAA10

AAAA

11AAAA12

AAA

13AAA14

AAAA

15AA A16

WT Q68A R69A P71A D72A Y73A N.C.

Gpr124 *SU����ǻLRR *SU����ǻIg *SU����ǻLRR+Ig

*SU����ǻHormR *SU����ǻGAIN Gpr125 No Txn

Tra

nsf

ect

ed

CO

S-7

cells

F

0

1

2

3

4

5

50 100 150 2000

Time (min)

A6

20

nm

Reck(CC1-5)

AlanineScanning

Mutagenesis

Fc Baits

COMP-APProbes

Gpr124(LRR-Ig)

Reck(CC1)Reck(CC1-2)Reck(CC1-5)

Reck(CC1-5) COMP-AP ProbeReck(CC1)Reck(CC1-2)Reck(CC1-5)

G 40 50 60 70 80 90 100

WT 2 7 9 10 14 15 16 N.C.

I

Reck 9 10 9+10 noReck

20

0

40

60

80

100

Re

lativ

e lu

cife

rase

act

ivity

C

Rec

k ǻ

ex2/ǻ

ex2 D;

Pdg

fb-C

reE

R

Rec

k ǻ

ex2/ǻ

ex2

Ctn

nb1

flex3

/+;

GS Lectin PECAM Bright-field GS Lectin PECAM Bright-field GS Lectin PECAM

Reck +/ǻex2; +/ǻGpr124

Reck Cr/ǻex2 ; +/ǻGpr124

GS Lectin PECAM GS Lectin PECAMJFc bait: Gpr124(LRR-Ig)

Reck CC1 Ala Scan COMP-AP probes

Figure 1. The Reck-Gpr124 complex transduces Wnt7a- and Wnt7b-specific signals to promote CNS angiogenesis. (A) Comparison of 19 Wnt ligands and Norrin transfected with or without Gpr124 and/or Reck in STF reporter cells. Arrows point to Wnt7a and Wnt7b synergistic responses. (B-D) E13.5 embryos are shown alongside coronal sections of the medial ganglionic eminence (MGE). GS Lectin and PECAM are endothelial markers. GS Lectin also stains for macrophages (green punctae). (C) Reck ex2/ ex2 mutant embryos show hemorrhaging in the forebrain (arrow) and distal spinal cord (arrowhead) with reduced vascular density and increased macrophage infiltration in the MGE. (D) -catenin was artificially stabilized in Reck ex2/ ex2; Ctnnb1flex3/+; Pdgfb-CreER embryos, which exhibit little or no hemorrhaging in the forebrain and spinal cord and show a near-complete rescue of CNS vascularization. (E) AP-binding assay using COS-7 cells transfected with Gpr124, Gpr124 deletion mutants, or Gpr125, a close homolog of Gpr124. Live cells were incubated with the Reck(CC1-5) COMP-AP probe. No Txn, no transfection. (F) Cell-free binding assay using Gpr124(LRR-Ig) or Reck(CC1-5) (a negative control) Fc baits captured in protein-G-coated wells. Each bait was incubated with three COMP-AP probes: Reck(CC1), Reck(CC1-2), and Reck(CC1-5). After removing unbound probe, the bound probe was visualized with a colorimetric AP reaction. (G) Amino acid sequence of mouse Reck CC1, with alanine scanning mutants indicated. Alanine substitution blocks 2, 7, 9, 10, 14, 15, and 16 were produced at detectable levels. (H) Images of the AP reactions at the final time point of cell-free binding assays with the indicated alanine scanning mutants (top) and single alanine substitutions (bottom). (I) STF cells were transfected with Wnt7a, Gpr124, and Reck WT or Ala Scan mutant plasmids. (J) Mice carrying an allele with alanine block substitutions 9+10 were generated using CRISPR/Cas9. E13.5 coronal sections of the lateral ganglionic eminence (LGE) of Reck+/ ex2; Gpr124+/ (control) and ReckCr/ ex2; Gpr124+/ (CRISPR mutant) animals are shown. The mutant animal displays reduced vascular density and increased macrophage infiltration in the LGE, compared to control animals. Scale bars, 200 m.

A

C D

Frizzled

Lrp5/6 Gpr124

Wnt7a or Wnt7b

Reck

B

Tran

sfec

ted

HE

K29

3T c

ells

Gpr124+ Wnt7a

Gpr124+ Fz4

Gpr124 Fz4 Wnt7a + Fz4Wnt7a

Gpr124+ Wnt7a

+ Fz4

Gpr124+ Wnt7a

+ Fz4 + Lrp5

Reck(CC1-5)-AP Probe

+ Gpr124 + Fz4

Wnt1 Wnt2 Wnt2b Wnt3 Wnt3a

Wnt7b Wnt8a Wnt8b Wnt9a Wnt9b

Wnt10a Wnt10b Wnt11 Wnt16 Norrin

Wnt4 Wnt5a Wnt5b Wnt6 Wnt7a

Reck(CC1-5)-AP Probe

+ Reck + Fz4

Wnt1 Wnt2 Wnt2b Wnt3 Wnt3a

Wnt7b Wnt8a Wnt8b Wnt9a Wnt9b

Wnt10a Wnt10b Wnt11 Wnt16 Norrin

Wnt4 Wnt5a Wnt5b Wnt6 Wnt7a

Tran

sfec

ted

HE

K29

3T c

ells

AP-Gpr124(LRR-Ig) Probe

Tran

sfec

ted

HE

K29

3T c

ells

Fz1 Fz2 Fz3 Fz4 Fz5 Fz6

Fz7 Fz8 Fz9 Fz10 No Txn

Fz1 Fz2 Fz3 Fz4 Fz5 Fz6

Fz7 Fz8 Fz9 Fz10 No Txn

+ Wnt7a+ Gpr124

+ Wnt7b+ Gpr124

Reck(CC1-5)-AP Probe

Tran

sfec

ted

HE

K29

3T c

ells

Reck+ Wnt7a

Reck+ Fz4

Reck Fz4 Wnt7a + Fz4Wnt7a

Reck+ Wnt7a

+ Fz4

AP-Gpr124(LRR-Ig) Probe

Reck+ Wnt7a

+ Fz4 + Lrp5

Tran

sfec

ted

HE

K29

3T c

ells

Figure 2. Reck and Gpr124 assemble in a multi-protein complex with Wnt7a/Wnt7b and Frizzled. (A) AP binding assay using HEK293T cells transfected with Gpr124, Wnt7a, Fz4, and/or Lrp5 (left) or Reck, Wnt7a, Fz4, and/or Lrp5 (right). Live cells were incubated with Reck(CC1-5)-AP (left) or AP-Gpr124(LRR-Ig) (right). (B) AP binding assay using HEK293T cells transfected with Gpr124 and Fz4 (left) or Reck and Fz4 (right), along with each of the 19 Wnts and Norrin. Live cells were incubated with Reck(CC1-5)-AP (left) or AP-Gpr124(LRR-Ig) (right). (C) AP binding assay using HEK293T cells transfected with Gpr124 and Wnt7a (top) or Wnt7b (bottom), together with each of the 10 Fz receptors or a no transfection control. Live cells were incubated with Reck(CC1-5)-AP. (D) Model of the multi-protein signaling complex, consisting of Reck, Gpr124, Wnt7a or Wnt7b, Frizzled, and Lrp5 or Lrp6. The straight double-headed arrow indicates a direct interaction between Reck(CC1) and Gpr124(LRR-Ig). The curved double-headed arrow indicates a direct interaction between Reck/Gpr124 and Fz/(Wnt7a or Wnt7b) sub-complexes. Scale bars, 200 m.