POSTER SESSION - uu-pharmfaculty50years.seuu-pharmfaculty50years.se/wp-content/uploads/2018/10/Poster-session... · i FACULTY OF PHARMACY 50 YEARS POSTER SESSION 26 October 2018 Full

POSTER SESSION

POSTER SESSION

BOOK OF ABSTRACTS

Faculty of Pharmacy

50 Years

26 October 2018

POSTER SESSION

POSTER SESSION

BOOK OF ABSTRACTS

Faculty of Pharmacy

50 Years

26 October 2018

Current research at the Faculty of Pharmacy is in focus during the poster session.

The 58 posters on display demonstrate the multidisciplinary aspects of pharmaceutical

sciences.

Posters are presented by the new generation of scientists.

the

Anniversary Organisation Committee and the Dean of The Faculty of Pharmacy

i

FACULTY OF PHARMACY 50 YEARS POSTER SESSION

26 October 2018

Full abstracts on QR:

1. THE ADDUCT FORMATION IN HILIC-ESI-MS IS STRONGLY AFFECTED BY THE INORGANIC ION CONCENTRATION OF THE SAMPLES. I. Erngren, J. Haglöf, M. Engskog, T. Arvidsson, C. Pettersson. [email protected] 2. IN VITRO LIPOLYSIS-PERMEATION ASSAY REFLECTS FENOFIBRATE ABSORPTION FROM LIPID-BASED FORMULATIONS IN PIGS. Janneke Keemink, Elin Mårtensson, Christel A.S. Bergström. [email protected] 3. SYNTHESIS OF SULFONIMIDAMIDE (SIA) BASED AMINO ACID BUILDING BLOCKS. Praveen Kumar Chinthakindi, Andrea Benediktsdottir, Edouard Zamaratski, Per I Arvidsson, Yantao Chen, and Anja Sandström. [email protected] 4. INTEGRATION OF CLINICAL PHARMACOKINETICS AND PHARMACODYNAMICS OF CLOFAZIMINE AND PYRAZINAMIDE TO A MULTISTATE TUBERCULOSIS PHARMACOMETRIC MODEL: EXPLAINS AN UNEXPECTED OUTCOME IN PHASE IIA TRIAL. Alan Faraj, Robin J Svensson, Ulrika SH Simonsson. [email protected] 5. ASYMMETRIC HINDLIMB MOTOR RESPONSE TO FOCAL TRAUMATIC BRAIN INJURY IS CONTROLLED BY SIDE-SPECIFIC OPIOID MECHANISM. H. Watanabe, M. Zhang (b,c), D. Sarkisyan, O. Kononenko, F. Clausen, T. Iakovleva, N. Marklund and G. Bakalkin. [email protected] 6. ADVANCED BIOMARKER DISCOVERY THROUGH INVESTIGATION OF GUT MICROBIOTA AND HUMAN HOST CO-METABOLISM - LINKING METABOLOMICS WITH CHEMICAL BIOLOGY METHODOLOGIES. Mário S. P. Correia, Caroline Ballet, Louis P. Conway, Neeraj Garg, Daniel Globisch (a, b). [email protected] 7. DEEP-SEA, DEEP SEQUENCING: WGS OF THE DEMOSPONGE GEODIA BARRETTI. Karin Steffen and Paco Cárdenas. [email protected] 8. BORONIC ESTER MACROCYCLES AS NEW E.COLI TYPE I SIGNAL PEPTIDASE INHIBITORS. Natalia Szałaj, Lu Lu, Andrea Benediktsdottir, Edouard Zamaratski, Sha Cao c, Gustav Olanders, Charles Hedgecock, Anders Karlén, Máté Erdélyi, Diarmaid Hughes, Sherry L. Mowbray, Peter Brandt. [email protected] 9. INVESTIGATION OF PULMONARY DISSOLUTION OF LOW SOLUBILITY DRUG USING AN EX VIVO LUNG MODEL AND COMPUTATIONAL ANALYSIS. J. Eriksson, H. Thörn, E. Sjögren, L. Holmstén, K. Rubin, H. Lennernäs. [email protected] 10. LANSOPRAZOLE PRE-TREATMENT INCREASES CYTOTOXICITY OF NANOFORMULATION (DOXIL) IN 2D CELL MODEL. F. Kullenberg (a,), O. Degerstedt, N. Pavlovic, F. Heindryckx, and H. Lennernäs. [email protected] 11. A MICROSERVICE-BASED INFRASTRUCTURE FOR LIGAND-BASED MODELING. Jonathan Alvarsson, B Ernst Ahlberg Helgee, A. Arvid Berg, A. Ola Spjuth. [email protected] 12. A HIGH-THROUGHPUT METHOD TO PREDICT THE IN VIVO EXPOSURE OF DIFFERENT DRUG DELIVERY SYSTEMS. C. Alvebratt, J. Keemink, K. Edueng, O. Cheung, M. Strømme, C. A.S. Bergström. [email protected] 13. SYNERGY CONFORMAL PREDICTION. Niharika Gauraha, Ola Spjuth. [email protected] 14. ISOLATION AND IDENTIFICATION OF THE MAIN STEROL FROM THE SPONGE HALICLONA SP. FROM TAIWAN. Merran Dunford (a,b), Sanjeevan Ranjendran, Patric Jonsson, Ping-Jyun Sung, Fang-Rong Chang, Achyut Adhikari, Sunithi Gunasekera, Paco Cárdenas. [email protected]

15. A DIGESTION/PERMEATION MODEL BASED ON ARTIFICIAL MEMBRANES FOR STUDIES OF PERFORMANCE OF LIPID-BASED FORMULATIONS. O. J. Hedge, C. A. S. Bergström. [email protected] 16. MAPPING THE DISTRIBUTION OF NEUROPEPTIDES THROUGH MASS SPECTROMETRY IMAGING. Heather E Hulme a, Halla Gunnarsdóttir a, Elva Fridjonsdottir a, Anna Nilsson a, Mohammadreza Shariatgorji a, Erwan Bezard b, Per E. Andrén a. [email protected] 17. DOCKING OF MACROCYCLES: COMPARING RIGID AND FLEXIBLE DOCKING IN GLIDE. Hiba Alogheli, Gustav Olanders, Wesley Schaal, Peter Brandt, Anders Karlén. [email protected] 18. HOME VS. SELF-INITIATED ART REFILL: CLINICAL, IMMUNOLOGICAL, AND VIROLOGIC OUTCOMES. Rory F. Leisegang, Jane Ball, Mark Cotton, David Dowdy, Keri Calkins, Susan Cleary, Gary Maartens, Jean B. Nachega. [email protected] 19. CHEMOSELECTIVE PROBES FOR THE DETECTION OF MICROBIAL BIOMARKERS OF DISEASE - COMBINING METABOLOMICS WITH CHEMICAL BIOLOGY METHODOLOGIES. Neeraj Garg, Louis P. Conway, Caroline Ballet, Mário S. P. Correia, Daniel Globisch.. [email protected] 20. LCMS-BASED UNTARGETED METABOLOMICS USING GUINEA PIG PERILYMPH: HIGHLIGHTING DIFFERENCES IN METABOLOME WITH AND WITHOUT A POSTERIORI HYDROGEN GAS ADMINISTRATION FOLLOWING LOUD NOISE EXPOSURE. K. Pirttilä, A. E. Fransson, J. Haglöf, M. Engskog, P. V. Pierre, G. Laurell, T. Arvidsson (a, d), C. Pettersson. [email protected] 21. POPULATION MODELING OF UNI-AND THREE-DIMENSIONAL AND DENSITY-BASED TUMOR MEASUREMENTS IN GASTRO-INTESTINAL STROMAL TUMOR PATIENTS TREATED WITH IMATINIB. Sreenath M. Krishnan, Emilie Schindler, Gaia Schiavon (b,c), Ron Mathijssen, Lena E. Friberg. [email protected] 22. POPULATION PHARMACOKINETIC ANALYSIS OF FACTOR VIII ACTIVITY FOLLOWING TREATMENT WITH MOROCTOCOG ALFA IN MODERATE TO SEVERE HAEMOPHILIA A SUBJECTS. João A. Abrantes, Elisabet I. Nielsen, Joan Korth-Bradley, Lutz Harnisch, Siv Jönsson. [email protected] 23. PROBABILISTIC PREDICTIONS OF METABOLISM USING VENN-PREDICTORS. Staffan Arvidsson Mc Shane, Lars Carlsson, Paolo Toccaceli, Ola Spjuth. [email protected] 24. CONTRALESIONAL HINDLIMB MOTOR RESPONSE INDUCED BY UNILATERAL BRAIN INJURY: EVIDENCE FOR EXTRA- SPINAL MECHANISM. N. Lukoyanov, L. Carvalho, H. Watanabe, M. Zhang (c, d), D. Sarkisyan, O. Kononenko, I. Bazov, T. Iakovleva, J. Schouenborg and G. Bakalkin. [email protected] 25. EXPLORING THE USEFULNESS OF MORPHOLOGICAL PROFILING OF CELLS TO STUDY TOXICITY MECHANISMS. P. Georgieva, Tanya Aggarwal, W. Schaal, O. Spjuth. [email protected] 26. EXPLORING THE USEFULNESS OF MORPHOLOGICAL PROFILING OF CELLS TO STUDY TOXICITY MECHANISMS. P. Georgieva, T. Aggarwal, W. Schaal, O. Spjuth. [email protected] 27. DISCREPANCY IN STABLY EXPRESSED PROTEINS WITHIN AND ACROSS HUMAN TISSUE TYPES. C. Wegler (a,b), M. Ölander, P. Artursson. [email protected] 28. EARLY-LIFE EXPOSURE TO TO AN ALGAL NEUROTOXIN INDUCES LONG-TERM CEREBELLAR NEURODEGENERATION AND LYSOSOMAL CHANGES. J Swan, L Ersson, A-L Berg, O Karlsson(a,c), E Brittebo. [email protected] 29. MASS SPECTROMETRY LIPIDOMICS: A PLATFORM TO UNRAVEL LIPID BIOLOGY. David Balgoma, Alfhild Grönbladh, Sofia Zelleroth, Mathias Hallberg, Curt Petterson, Mikael Hedeland. [email protected] 30. DISCOVERY AND CHARACTERIZATION OF SMALL PEPTIDE TOXINS FROM THE EPIDERMAL MUCUS OF LINEUS LONGISSIMUS, THE LONGEST ANIMAL ON EARTH. Erik Jacobsson, Håkan S. Andersson, Malin Strand, Steve Peigneur, Eline K. M. Lebbe, K. Johan Rosengren, Jan Tytgat , Ulf Göransson. [email protected] 31. CLINICAL METABOLOMICS: THE STETHOSCOPE FOR THE 21ST CENTURY? Stephanie Hermana (a,b), Payam Emami Khoonsari, Valter Niemelä, Torbjörn Åkerfeldta, Ola

ii

Spjuth, Joachim Burman, Kim Kultima. [email protected] 32. SINGLE BEAD INVESTIGATION OF A CLINICAL DRUG DELIVERY SYSTEM. E. Ahnfelt, J. Gernandt, Y. Al-Tikriti, E. Sjögren, H. Lennernäs, and P. Hansson. [email protected] 33. CLASSIFICATION OF DRUG INDUCED TOXICOLOGY IN TISSUE SAMPLES USING MALDI MASS SPECTROMETRY IMAGING. A. Nilsson, R.J. Goodwin, B.Forngren, S. Iverson, J. Lindberg, P.E. Andrén. [email protected] 34. SCILIFELAB DRUG DISCOVERY AND DEVELOPMENT PLATFORM- MEDICINAL CHEMISTRY. U. Yngve and Y. Gravenfors. [email protected] 35. ELEVATED BRAIN LEVELS OF L-DOPA AND STRUCTURE SPECIFIC CHANGES IN DOPAMINE METABOLISM IN L-DOPA INDUCED DYSKINESIA REVEALED BY MASS SPECTROMETRY IMAGING. Elva Friðjónsdóttir, Mohammadreza Shariatorji, Anna Nilsson, Theodosia Vallianatou, Per Svenningsson, Erwan Bezard, Per E. Andrén. [email protected] 36. A MASS SPECTROMETRY IMAGING APPROACH FOR INVESTIGATING HOW DRUG-DRUG INTERACTIONS INFLUENCE DRUG BLOOD-BRAIN BARRIER PERMEABILITY. Theodosia Vallianatou, Nicole Strittmatter, Anna Nilsson, Mohammadreza Shariatgorji, Gregory Hamm, Marcela Pereira, Patrik Källback, Per Svenningsson, Maria Karlgren, Richard J.A. Goodwin, A. Per E. Andrén. [email protected] 37. ANABOLIC ANDROGENIC STEROIDS INDUCE TOXIC EFFECTS IN RAT PRIMARY CORTICAL CELL CULTURES. Sofia Zelleroth, Erik Nylander, Fred Nyberg, Alfhild Grönbladh, Mathias Hallberg. [email protected] 38. CAUSAL FACTORS FOR POST-CRYOPRESERVATION VARIABILITY IN ATTACHMENT AND CULTURE MORPHOLOGY OF PRIMARY HUMAN HEPATOCYTES. M. Ölander, J.R. Wisniewski, I. Flörkemeier, A. Norén, and P. Artursson. [email protected] 39. COMPLETE KNOCKOUT OF ENDOGENOUS MDR1 (ABCB1) IN MDCK WILDTYPE AND MDCK-HMDR1 CELLS BY CRISPR-CAS9. M. Karlgren (a, b), I. Simoff, M. Backlund (a, b), C. Wegler (a, c), M. Keiser, N. Handin, J. Müller, P. Lundquist, A-C. Jareborg, S. Oswald, P. Artursson (a, b). [email protected] 40. TARGETED BRAIN DELIVERY OF METHOTREXATE BY GLUTATHIONE PEGYLATED LIPOSOMES: HOW CAN THE FORMULATION MAKE A DIFFERENCE? Yang Hu, Pieter J. Gaillard, Elizabeth C.M. de Lange, and Margareta Hammarlund-Udenaes. [email protected] 41. MAPPING THE BRAIN NEUROTRANSMITTER NETWORK WITH MALDI MASS SPECTROMETRY IMAGING. M. Shariatgorji, A. Nilsson, L. Odell, E. Fridjonsdottir, T. Vallianatou, L. Katan, J. Sävmarker, E. Bezard, P. Svenningsson, P. E. Andren. [email protected] 42. THE INNATE/ADAPTIVE IMMUNE RESPONSE TRIGGERED UPON LOCAL IMMUNOTHERAPY OF ORTHOTOPICALLY GROWING BLADDER CANCER TUMORS. Iliana K. Kerzeli, Andrea Kurtinovic, Aikaterini Nasi, Ivan Stepánek, Martin Lord, Sara M. Mangsbo. [email protected] 43. THE UPPSALA UNIVERSITY DRUG OPTIMIZATION AND PHARMACEUTICAL PROFILING FACILITY (UDOPP) — EMERGING NATIONAL INFRASTRUCTURE FOR ADME INVESTIGATIONS IN SWEDEN. Pawel Baranczewski, Maria Backlund, Maria Karlgren, Shibu Krishnan, Annika Lindqvist, Maria Mastej, Aljona Saleh, Ivailo Simoff, Richard Svensson and Per Artursson. [email protected] 44. METABOLIC CONTROL OF DENDRITIC CELL IMMUNOGENICITY. Aikaterini Nasi (a.b), Vishnu Priya Bollampalli, Meng Sun, Yang Chen, Sylvie Amu, Susanne Nylén, Liv Eidsmo, Antonio Gigliotti Rothfuchs, Bence Réthi. [email protected] 45. STABILIZED CYCLIC PEPTIDES AS SCAVENGERS OF AUTOANTIBODIES: NEUTRALIZATION OF ANTI-CITRULLINATED PROTEIN/PEPTIDE ANTIBODIES IN RHEUMATOID ARTHRITIS. Sunithi Gunasekera, Cátia Fernandes-Cerqueira, Stefan Wennmalm, Heidi Wähämaa, Yngve Sommarin, Anca I. Catrina, Per-Johan Jakobsson, Ulf Göransson. [email protected] 46. PHOSPHOLIPID BUT NOT NEUTRAL LIPID CONTENT IS A MAJOR DETERMINANT OF INTRACELLULAR BINDING OF DRUGS. Andrea Treyer,

André Mateus, Hinnerk Boriss, Pär Matsson, Per Artursson. [email protected] 47. THE PROTEOME OF PRIMARY HUMAN HEPATOCYTE SPHEROIDS CULTURED AND USED IN HIGH- THROUGHPUT SCREENING FORMAT. Niklas HANDIN, Wojciech SENKOWSKI, Magnus ÖLANDER, Frida NYBERG, Mårten FRYKNÄS, Per ARTURSSON. [email protected] 48. TRANSLATING LONG-READ SEQUENCING TO THE CLINIC: A CASE STUDY USING BCR-ABL1 MUTATION SCREENING IN CHRONIC MYELOID LEUKEMIA. Wesley Schaal, Adam Ameur (b,c,) Ulla Olsson-Strömberg, Lucia Cavelier, Ola Spjuth (a,c). [email protected] 49. PALLADIUM(0)-CATALYZED CARBONYLATIVE SYNTHESIS OF N-ACYLSULFONAMIDES USING MOLYBDENUM HEXACARBONYL AS A SOLID CARBON MONOXIDE SOURCE. Luke S. Schembri, Luke R. Odell. [email protected] 50. TRANSFORMATION OF KR-12 DERIVED CROSS-LINKED CYCLIC DIMERS INTO STABLE AND POTENT ANTIMICROBIAL DRUG LEADS. Taj Muhammad, Adam A. Strömstedt,, Sunithi Gunasekera, Ulf Göransson. [email protected] 51. TIME-DEPENDENT EFFECTS ON SMALL INTESTINAL TRANSPORT BY ABSORPTION-MODIFYING EXCIPIENTS. Dahlgren, David; Roos, Carl; Tannergren, Christer4; Lundqvist, Anders; Sjöblom, Markus; Sjögren, Erik; Lennernäs, Hans. [email protected] 52. EFFECT OF ABSORPTION-MODIFYING EXCIPIENTS, HYPOTONICITY, AND ENTERIC NEURAL ACTIVITY IN AN IN VIVO MODEL FOR SMALL INTESTINAL TRANSPORT. Dahlgren, David1; Roos, Carl; Tannergren, Christer; Lundqvist, Anders; Sjöblom, Markus; Sjögren, Erik; Lennernäs, Hans. [email protected] 53. THE EFFECTS OF THREE ABSORPTION-MODIFYING CRITICAL EXCIPIENTS ON THE IN VIVO INTESTINAL ABSORPTION OF SIX MODEL COMPOUNDS IN RATS AND DOGS. Dahlgren, David; Roos, Carl; Johansson, Pernilla; Tannergren, Christer; Lundqvist, Anders; Langguth, Peter; Sjöblom, Markus; Sjögren, Erik; Lennernäs, Hans. [email protected] 54. PRECLINICAL EFFECT OF ABSORPTION-MODIFYING EXCIPIENTS ON RAT INTESTINAL TRANSPORT OF MODEL COMPOUNDS AND THE MUCOSAL BARRIER MARKER 51CR-EDTA. Dahlgren, David; Roos, Carl; Tannergren, Christer; Lundqvist, Anders; Langguth, Peter; Sjöblom, Markus; Sjögren, Erik; Lennernäs, Hans. [email protected] 55. TRANSFER LEARNING WITH DEEP CONVOLUTIONAL NEURAL NETWORKS FOR CLASSIFYING CELLULAR MORPHOLOGICAL CHANGES. Alexander Kensert, Philip J Harrison, Ola Spjuth. [email protected] 56. SUCCESSFUL METHOD VALIDATION AND ABSOLUTE QUANTITATION BY MALDI MASS SPECTROMETRY IMAGING. Patrik Källback, Theodosia Vallianatou, Anna Nilsson, Mohammadreza Shariatgorji, Nicoletta Schintu, Marcela Pereira, Florian Barré, Henrik Wadensten, Per Svenningsson and Per E. Andrén. [email protected] 57. JEJUNAL ABSORPTION OF APREPITANT FROM SUSPENSIONSIN RAT: ROLE OF PARTICLE SIZE, PRANDIAL STATE AND MUCUS LAYER. Carl Roos, David Dahlgren, Erik Sjögren, Markus Sjöblom, Mikael Hedeland, Hans Lennernäs. [email protected] 58. MECHANISTIC MODELLING OF INTESTINAL DRUG ABSORPTION – THE IN VIVO EFFECTS OF NANOPARTICLES, HYDRODYNAMICS, AND COLLOIDAL STRUCTURES. Carl Roos, Jan Westergren, David Dahlgren, Hans Lennernäs, Erik Sjögren. [email protected] 59. DEVELOPMENT OF SCAFFOLD-BASED PEPTIDE SCAVENGERS FOR INHIBITION OF AUTOANTIBODIES IN RHEUMATOID ARTHRITIS. Camilla Eriksson, Sunithi Gunasekera, Cátia Fernandes-Cerqueira, Per-Johan Jakobsson, Ulf Göransson. [email protected] HTTP://UU-PHARMFACULTY50YEARS.SE/WP-CONTENT/UPLOADS/2018/10/POSTER-SESSION-BOOK-OF-ABSTRACT.PDF

Poster session Faculty of Pharmacy 50 Year Anniversary

1

1

THE ADDUCT FORMATION IN HILIC-ESI-MS IS STRONGLY AFFECTED BY THE INORGANIC ION CONCENTRATION OF THE SAMPLES

I. Erngren (a), J. Haglöf (a), M. Engskog (a), T. Arvidsson (a,b), C. Pettersson (a)

[email protected]

(a) Department of Medicinal Chemistry, Uppsala University, (b) Medical Product Agency, Uppsala

Introduction. In metabolomics electrospray-mass spectrometry (ESI-MS) hyphenated with liquid chromatography (LC) is one of the most commonly used techniques. However, in ESI-MS adduct formation is very common and often seen as a major pitfall which lead to complicated data sets and sensitivity loss where metabolites are represented by several different adducts. Adding to this complexity, inorganic ions such as sodium and potassium show retention similar to many metabolites in hydrophilic interaction liquid chromatography (HILIC) leading to co-elution of metabolites and inorganic ions which could affect the adduct formation of these metabolites. The correlation between metabolite adduct formation and co-elution with sodium and potassium was studied by looking at trends between adduct formation and retention of metabolites in cell-samples analysed by HILIC-ESI-QTOF-MS. Methods. Head and neck cancer cells were cultivated and harvested, the intracellular metabolome was extracted using chloroform:methanol:water 4:4:2.85. The aqueous phase was isolated and analysed using an Acquity UPLC I-class system from Waters (Manchester, UK) coupled to a G2S Synapt Q-TOF equipped with an electro spray ionization (ESI) source (Waters). The separation was performed on an Acquity UPLC BEH Amide Column (1.7 µM, 2.1x50 mm) from Waters and the detection was performed in positive ionization mode using data independent acquisition, MSE. The adduct formation of forty annotated metabolites was investigated, correlation between the types of adducts, retention time and co-elution with sodium and potassium formate clusters was studied. Preliminary data. It was found that sodium and potassium adducts were formed in much higher abundance for those metabolites that co-eluted with sodium and potassium formate clusters. In the retention time periods of sodium cluster (RT=8-9 min) and potassium cluster (RT= 7-8 min) almost all metabolites were detected as sodium or potassium adducts respectively. To test if the adduct formation was a result of the different metabolites structures; since both retention and adduct formation is dependent upon molecule structure or if it in fact was the co-elution of metabolites and alkali metal ions that affected the adduct formation, post-column infusion experiments were carried out. Ten metabolites were infused post-column while standard solutions with varying concentrations of sodium and potassium formate (1 - 100mM) was injected. The adduct formation of the infused metabolites was tracked during the elution of sodium and potassium. It was found that the adduct formation of the infused metabolites were strongly affected by the elution of sodium and potassium. The enhancement of the sodium and potassium adducts were as high as 8×106% as compared to a solvent blank injection for some metabolites such as phenylalanine and guanosine. While the suppression of the [M+H] adduct was as high as 97% as compared to baseline in the same samples. The enhancements of the sodium and potassium adducts were not increased linearly with increasing sodium and potassium concentration. In an analytical point of view this renders several issues that need to be addressed, in relative quantitation where no internal standards are used, there is an obvious risk to use the adduct signals for quantitation since the differences might be a secondary effect of different levels of alkali metal ions in the samples. The adduct formation also yields numerous signals from each metabolite making data interpretation much more complicated. Novel aspect. The outcome of relative quantitations in MS could be strongly influenced by the sample alkali metal ion content.


2

2

IN VITRO LIPOLYSIS-PERMEATION ASSAY REFLECTS FENOFIBRATE ABSORPTION FROM LIPID-BASED FORMULATIONS IN PIGS

Janneke Keemink, Elin Mårtensson, Christel A.S. Bergström

[email protected]

Department of Pharmacy, Uppsala University, Uppsala Biomedical Center, P.O. Box 580, SE-751 23 Uppsala, Sweden

Purpose. Lipid-based formulations (LBFs) are a delivery strategy to enhance intestinal absorption of poorly soluble compounds by keeping them solubilized during their transit in the gastrointestinal tract [1]. Unfortunately, in vitro lipolysis studies, typically used to evaluate performance of LBFs, do not accurately predict the in vivo behavior of LBFs [2]. One of the main reasons for this poor correlation is the absence of an absorption compartment in the in vitro lipolysis setup [1]. To better predict in vivo drug absorption from LBFs, we developed an in vitro method that simultaneously evaluates digestion and absorption. Methods. In vitro lipolysis was carried out in a lipolysis-permeation setup consisting of two chambers, representing a digestion and an absorption chamber, separated by a Caco-2 monolayer. Digestions of three fenofibrate-loaded-LBFs were performed with immobilized lipase. Samples were collected from both chambers to determine (i) fenofibrate solubility in the aqueous phase of the digestion medium and (ii) fenofibrate permeation across the Caco-2 monolayer. In vivo data from the literature were used to perform an in vitro-in vivo correlation [2]. Results and Discussion. Fenofibrate concentrations in the aqueous phase of the digestion medium predicted significantly higher exposure upon administration of type IIIA formulations than of the type IV formulation. However, mass transfer of fenofibrate was similar for the three formulations, thereby indicating that in vivo exposure of fenofibrate would also be similar. Conclusions. Drug transfer across the Caco-2 membrane accurately reflected in vivo drug exposure upon administration of three different LBFs loaded with fenofibrate. Acknowledgements. This work has received support from the European Research Council Grant 638965 and the Swedish Research Council Grant 2014-3309. References [1] O.M. Feeney et al. Adv. Drug Deliv. Rev. 101 (2016), [2] B.T. Griffin et al. Eur. J. Pharm. Biopharm. EV. 86 (2014) 427�437.


3

3

SYNTHESIS OF SULFONIMIDAMIDE (SIA) BASED AMINO ACID BUILDING BLOCKS

Praveen Kumar Chinthakindi (a), Andrea Benediktsdottir (a), Edouard Zamaratski (a), Per I Arvidsson

(c), Yantao Chen (b), and Anja Sandström (a) [email protected]

(a) Department of Medicinal Chemistry, Drug Design and Discovery, BMC, Uppsala University, Sweden. (b) Cardiovascular and Metabolic Diseases, IMED, AstraZeneca R&D, Gothenburg, Sweden(c) Medical

Biochemistry and Biophysics, Karolinska Institutet-Scilife Lab, Stockholm, Sweden Amide bond isosteres and unnatural amino acids are important building blocks in the development of pseudopeptides and peptidomimetics.[1] They can modulate a peptide physiochemical propertiesas well as selectivity against biological targets. Expanding on the recent tactical application of bioisosteres[2] i.e., sulfur-aza class of analogues,[3] Arvidsson group recently highlighted an emerging interest in sulfonimidamides (SIA) in contemporary drug design.[4] SIA is a chiral isostere of sulfonamide that offers number of advantages in drug design, for instance building highquality compound libraries.[5] SIAs have an extra "N" atom/handle in comparison with sulfonamide and imine N-substituents of sulfonimidamide are known to tune the physicochemical and biological properties.[6,7] Furthermore, in some cases single atom modifications are known to increase increase the activity of the parent molecule.[8] Therefore, the full potential of sulfonimidamides in regard to stability as well as biological activity should be further exploited in biologically relevant molecules. Acknowledgements. Anja Sandström gratefully acknowledges Kjell and Märta Beijer Foundation for financial support. P. K. Chinthakindi acknowledges the Department of Medicinal Chemistry, Uppsala University, Sweden, for a postdoc fellowship. The authors wish to thank master students Ayah Ibrahim and Ataa Wared for their contributions. References: 1) E. Valeur, S. M. Guéret, H. Adihou, R. Gopalakrishnan, M. Lemurell, H. Waldmann, T. N. Grossmann, A. T. Plowright, Angew. Chemie Int. Ed. 2017, 56, 10294�10323. 2) N. A. Meanwell, J. Med. Chem. 2011, 54, 2529�2591. 3) U. Lücking, Angew. Chemie Int. Ed. 2013, 52, 9399�9408. 4) P. K. Chinthakindi, T. Naicker, N. Thota, T. Govender, H. G. Kruger, P. I. Arvidsson, Angew. Chemie Int. Ed. 2017, 56, 4100�4109. 5) F. W. Goldberg, J. G. Kettle, T. Kogej, M. W. D. Perry, N. P. Tomkinson, Drug Discov. Today 2015, 20, 11�17. 6) L. D. Pennington, D. T. Moustakas, J. Med. Chem. 2017, 60, 3552�3579. 7) S. R. Borhade, R. Svensson, P. Brandt, P. Artursson, P. I. Arvidsson, A. Sandström, ChemMedChem 2015, 10, 455�460. 8) D. L. Boger, J. Org. Chem. 2017, 82, 11961�11980.


4

4

INTEGRATION OF CLINICAL PHARMACOKINETICS AND PHARMACODYNAMICS OF CLOFAZIMINE AND PYRAZINAMIDE TO A MULTISTATE TUBERCULOSIS

PHARMACOMETRIC MODEL: EXPLAINS AN UNEXPECTED OUTCOME IN PHASE IIA TRIAL

Alan Faraj, Robin J Svensson, Ulrika SH Simonsson

[email protected]

Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden Tuberculosis (TB) is the main cause of death, within infectious diseases. In a Phase IIa clinical trial aiming to scrutinize the early bactericidal activity (EBA) of clofazimine (CFZ) and pyrazinamide (PZA) monotherapy, no significant drug effects was observed. Despite being an important part of the sterilizing multi-drug standard treatment of tuberculosis, a paradoxical numeric increase of bacterial load was observed, post CFZ treatment. Using non-linear mixed effects modeling, we characterized statistically significant exposure-response relationships of both drugs, and discovered a quantitative model-based explanation for the paradoxical increase in colony forming units (CFU) after CFZ treatment. Simulations based on the final model suggested that a linearly concentration-dependent effect on dormant tubercular bacilli, explains the increase in utilized bacterial biomarker. Sensitive analysis tools are desired to reduce impact of noisy data, and to accurately explore drug effects to advance clinical trials. We propose that this quantitative semi-mechanistic approach to study drug effects, provides a rational framework for analysing Phase IIa EBA studies, and can accelerate anti-TB drug development. Acknowledgements. We express gratitude towards all patients that participated in the clinical trial, as well as the principal investigators and staff that executed the trial. We thank the authors of the original publication for sharing data in accordance with GDPR, and enabling this interdisciplinary work. Contributions: A.Faraj developed the PK/PD model and initiated the writing of the manuscript. AH Diacon was the principal investigator of the clinical trial and provided clinical expertise regarding scientific plausibility of modeling results. USH Simonsson was the principal expert in the modeling and simulation analysis work, providing quality assurance in the data analysis. RJ Svensson provided hands on support and supervised A. Faraj during the master thesis work that resulted in this work. 1. A. H. Diacon, R. Dawson, F. von Groote-Bidlingmaier, G. Symons, A. Venter, P. R. Donald, C. van Niekerk, D. Everitt, J. Hutchings, D. A. Burger, R. Schall, C. M. Mendel, Bactericidal activity of pyrazinamide and clofazimine alone and in combinations with pretomanid and bedaquiline, Am. J. Respir. Crit. Care Med. 191, 943–953 (2015). 2. R. J. Svensson, S. H. Gillespie, U. S. H. Simonsson, Improved power for TB Phase IIa trials using a model-based pharmacokinetic-pharmacodynamic approach compared with commonly used analysis methods, J. Antimicrob. Chemother. 72, 2311–2319 (2017). 3. R. Svensson, U. Simonsson, Application of the Multistate Tuberculosis Pharmacometric Model in

Patients With Rifampicin‐Treated Pulmonary Tuberculosis, CPT Pharmacomet. Syst. Pharmacol. 5, 264–273 (2016). 4. O. Clewe, L. Aulin, Y. Hu, A. R. M. Coates, U. S. H. Simonsson, A multistate tuberculosis pharmacometric model: a framework for studying anti-tubercular drug effects in vitro, J. Antimicrob. Chemother. 71, 964–974 (2016).


5

5

ASYMMETRIC HINDLIMB MOTOR RESPONSE TO FOCAL TRAUMATIC BRAIN INJURY IS CONTROLLED BY SIDE-SPECIFIC OPIOID MECHANISM

H. Watanabe (a), M. Zhang (b,c), D. Sarkisyan (a), O. Kononenko (a), F. Clausen (d), T. Iakovleva (a),

N. Marklund (d) and G. Bakalkin (a) [email protected]

(a) Dept. Pharmaceutical Biosciences, Uppsala University, Sweden, (b) Neuronano Research Center, Lund University, Sweden, (c) Dept. Molecular Medicine, University of Southern Denmark, Odense,

Denmark, (d) Dept. Neuroscience, Uppsala University, Sweden

Neuropeptides exert specific, coherent effects on formation and rewiring of neural circuits and, consequently, on behavior. Most neuropeptides are expressed in tiny neuronal subpopulations that are characterized by specific distribution patterns. We examined whether these patterns are flexible and maybe changed upon transition to a new physiologic state. For this purpose, controlled cortical impact injury (CCI) and sham operation (SO) as a control were applied to perturb endogenous opioid system (EOS) brain and spinal circuits. CCI produces generalized effects that result in impairment of multiple processes, including motor and sensory functions, emotions, and cognition, all of which are regulated by the EOS. The striking feature of this system is its asymmetric coexpression patterns, which suggest side-specific regulation of selective neural circuits by opioid neurohormone. The controlled cortical impact injury-induced postural asymmetry is encoded at the spinal level. Opioid receptors mediate formation of the asymmetry and determine the flexion side. Pharmacological correction with opioid antagonists - a strategy to reverse motor deficits. Acknowledgements. Drs. DeAnna Adkins, MUSC; V. Galatenko, I. Mityakina, A. Tonevitsky, MSU, Moscow; Nikolay Lukoyanov, Universidade do Porto, Portugal. Supported by the Swedish Science Research Council


6

6

ADVANCED BIOMARKER DISCOVERY THROUGH INVESTIGATION OF GUT MICROBIOTA AND HUMAN HOST CO-METABOLISM -

LINKING METABOLOMICS WITH CHEMICAL BIOLOGY METHODOLOGIES

Mário S. P. Correia (a), Caroline Ballet (a), Louis P. Conway (a), Neeraj Garg (a), Daniel Globisch (a, b)

[email protected]

(a) Uppsala University, Uppsala, Sweden, (b) Science for Life Laboratories, Uppsala, Sweden The detailed investigation of metabolites in human samples (serum, plasma, urine, saliva, feces or tissues), termed metabolomics, carries a great potential for the discovery of unknown biomarkers. This is urgently required for diseases with advanced stage prognosis such as pancreatic cancer, one of the most lethal cancers worldwide with steadily increasing patient numbers. However, this is a challenging task as biological samples are comprised of a complex mixture of biomolecules. Advanced chemical biology tools are limited for this newest “omics”-research field. One of the most exciting scientific developments in the past decade has been the understanding that gut microbiota profoundly impact human physiology.1,2 The complex consortium of trillions of microbes possesses a wide range of metabolic activity. This metabolic interspecies communication represents a tremendous opportunity for biomarker discovery as only limited information on this co-metabolism has been elucidated on a molecular level.3 We have developed new chemical biology techniques for an enhanced metabolomics analysis using liquid chromatography-coupled with tandem mass spectrometry (UPLC-MS/MS). Our research is based on selective biochemical conversion of metabolites for qualitative and quantitative analysis through chemical synthesis of (isotope-labeled) reference molecules. We have combined these new methods with state-of-the-art mass spectrometric and bioinformatics techniques to enhance the scope of general metabolomics studies. Our approach represents a new and unique strategy for metabolite analysis and is focused on host-microbiota cometabolism. The detailed investigation of small molecule metabolites in human samples including feces, serum/plasma and urine, represents a high potential for discovery of unknown biomarkers. Application of our new metabolite analyzing methodologies at the interface of chemistry and biology led to the identification of a series of previously unknown metabolites using mass spectrometry-based metabolomics. (1) Turnbaugh, P. J. et al. A core gut microbiome in obese and lean twins. Nature 2009, 457 (7228), 480-4. (2) Liu, R. et al. Gut microbiome and serum metabolome alterations in obesity and after weight-lossintervention. Nat. Med. 2017, 23 (7), 859-868. (3) Wikoff, W. R et al Metabolomics analysis reveals large effects of gut microflora on mammalian blood metabolites. Proc. Natl. Acad. Sci. U. S. A. 2009, 106 (10), 3698-3703.


7

7

DEEP-SEA, DEEP SEQUENCING: WGS OF THE DEMOSPONGE GEODIA BARRETTI

Karin Steffen and Paco Cárdenas

[email protected]

Pharmacognosy, Department of Medicinal Chemistry, BMC, Uppsala University, Uppsala, Sweden

Sponges (phylum Porifera), among the oldest extant animals, are a poorly studied organismal group. However, the sponge fauna represents today one of the most diverse macrobenthic animal groups, more diverse than corals and probably as important as corals in the ecology of shallow and deep-sea ecosystems. Sponges also have one of the greatest potential in the rapidly expanding field of marine biotechnology and more specifically in the marine natural products research. With only a few whole genomes of shallow water species published so far, genomic data is truly lacking to investigate questions on deep-sea North Atlantic sponges. The aim of our project is therefore to sequence the whole genome of the deep-sea demosponge Geodia barretti (order Tetractinellida). With this new reference genome, we wish to answer questions on the sponge chemical potential and its natural products. In addition, this is also a unique opportunity to further elucidate early animal evolution.


8

8

BORONIC ESTER MACROCYCLES AS NEW E.COLI TYPE I SIGNAL PEPTIDASE INHIBITORS

Natalia Szałaj (a), Lu Lu (b), Andrea Benediktsdottir (a), Edouard Zamaratski (a), Sha Cao c, Gustav

Olanders (a), Charles Hedgecock (a), Anders Karlén (a), Máté Erdélyi (d), Diarmaid Hughes (c), Sherry L. Mowbray (b), Peter Brandt (a)

[email protected]

(a) Department of Medicinal Chemistry, Uppsala University, Box 574, SE-751 23 Uppsala, Sweden. b) Department of Cell and Molecular Biology, Uppsala University, Box 596, SE-751 24 Uppsala, Sweden. c) Department of Medical Biochemistry and Microbiology, Uppsala University, Box 582, SE-75 123 Uppsala,

Sweden. d) Department of Chemistry, BMC, Uppsala University, SE-751 23 Uppsala, Sweden.

Type I signal peptidase, with its vital role in bacterial viability, is a promising but underexploited antibacterial drug target. In the light of steadily increasing rates of antimicrobial resistance, we have developed novel macrocyclic lipopeptides, linking P2 and P1' by a boronic ester warhead, capable of inhibiting Escherichia coli type I signal peptidase (EcLepB) and exhibiting good antibacterial activity. Structural modifications of the macrocyclic ring, the peptide sequence and the lipophilic tail led us to 14 novel macrocyclic boronic esters. It could be shown that macrocyclization is well tolerated in terms of EcLepB inhibition and antibacterial activity. Among the synthesized macrocycles, potent enzyme inhibitors in the low nanomolar range (e.g. compound 42f, EcLepB IC50 = 29 nM) were identified also showing good antimicrobial activity (e.g. compound 42b, E. coli WT MIC = 16 µg/mL). The unique macrocyclic boronic esters described here were based on previously published linear lipopeptidic EcLepB inhibitors in an attempt to address cytotoxicity and hemolysis. We show herein that structural changes to the macrocyclic ring influence both the cytotoxicity and hemolytic activity suggesting that the P2 to P1' linker provide means for optimizing offtarget effects. However, for the present set of compounds we were not able to separate the antibacterial activity and cytotoxic effect.


9

9

INVESTIGATION OF PULMONARY DISSOLUTION OF LOW SOLUBILITY DRUG USING AN EX VIVO LUNG MODEL AND COMPUTATIONAL ANALYSIS

J. Eriksson (a), H. Thörn(b), E. Sjögren (a), L. Holmstén (b), K. Rubin (c), H. Lennernäs (a)

[email protected]

a) Department of Pharmacy, Uppsala University, Uppsala, 75233, Sweden, b) Pharmaceutical Technology & Development Inhalation, AstraZeneca, Pepparedsleden 1, Mölndal, 43183, Sweden, c) Respiratory,

Inflammation & Autoimmunity, Innovative Medicines and Early Development, AstraZeneca, Pepparedsleden 1, Mölndal, 431 83, Sweden

BACKGROUND: Drug dissolution in the lungs is an important process when considering pulmonary drug delivery. Among inhaled drug compounds, a large number have low solubility; consequently, dissolution is the likely rate-limiting step in the overall absorption process1. In vivo predictive models, such as the isolated perfused lung (IPL) model,2, can be used to improve the understanding of pulmonary dissolution of inhaled drug compounds. AIMS: Using the IPL model and computational analysis, the objective of this study was to improve the understanding of dissolution processes in the lungs of four selected low solubility compounds (candidate drug 1 (CD1), budesonide, fluticasone furoate and fluticasone propionate) administered as suspensions and dry powder formulations. RESULTS: The dissolution in the lungs was the rate-limiting absorption parameter for drug transport in the IPL model for all compounds except budesonide and the small particle size of CD1. This is probably due to the high solubility of budesonide and the small particle size of CD1 which generates a large surface area promoting the dissolution rate. The dry powder of the larger particle size of CD1 had a four times slower dissolution than the corresponding suspension. The estimated dissolution constant and the drug transport in the IPL model ranked accordingly to the solubility in PBS of the compounds and there was a good agreement between ex vivo and in vitro dissolution constant. CONCLUSIONS: The dissolution in the IPL model is affected by the size of the drug particles, the solubility of the compound and the formulation properties. In addition, the IPL model in combination with computational analysis has been shown to be a useful tool when investigating pulmonary dissolution and to increase the understanding of dissolution related parameters for inhaled drug compounds. REFERENCES: [1] May, S., et al. Pharm Res 31, 3211-3224(2014) [2] Eriksson, J., et al. Eur J Pharm Biopharm 124, 1-12(2018)


10

10

LANSOPRAZOLE PRE-TREATMENT INCREASES CYTOTOXICITY OF NANOFORMULATION (DOXIL) IN 2D CELL MODEL

F. Kullenberg (a,), O. Degerstedt (a), N. Pavlovic (b), F. Heindryckx (b), and H. Lennernäs (a)

[email protected]

(a) Department of Pharmacy, Uppsala University, Box 580, 751 23 Uppsala, Sweden (b) Department of Medical Cell Biology, Uppsala University, Box 571, Uppsala, Sweden

In recent years, countless nanoformulations of the cytostatic drug doxorubicin (DOX) have been developed with the aim of treating various recent cancers.[1] Doxil® (named Caelyx® in Europe), a liposomal formulation of DOX, was the first nanoformulation to receive regulatory approval in 1995 and is still one of the best example of successful clinical nanomedicines.[2] Much research has investigated the synergistic effects observed when giving a proton pump inhibitor (PPI) as a concomitant medication to conventional DOX treatment. Several clinical studies are ongoing and encouraging results have already been seen both in-vitro and in-vivo. [3] PPI’s appear to increase the amount of DOX located in the cell nucleus, by increasing extracellular pH and thereby reducing endosomal sequestration of the drug, resulting in higher cytotoxicity. [4] However, very little is known about how PPI pre-treatment will affect uptake and distribution of nanoformulations with DOX in tumor cells. The primary objective of this study was to compare the effect of pre-treating cancer cells with different concentrations of a common proton pump inhibitor (PPI) before exposure to Doxil® or DOX in solution. Though the objective was to compare the results in four different cell lines, this study was delayed due to contamination issues. As such, only the HCC cell line Huh7's results being shown in this poster. It is clear that the pre-treatment with 500 µM PPI improves the effect of this DOX nanoformulation by more than one order of magnitude (p<0.001). The effect on cell viability of DOX as a solution and in nanoformulation agree with earlier results5, which validates the experimental setup in our laboratory. The synergistic effect of our lansoprazole pre-treatment and a nanoformulation of DOX is clearly demonstrated in this in vitro study. There are several potential mechanisms that could explain this observation, as shown in the figure above. These mechanisms and their effect on cytotoxicity will be investigated in future studies in different cell lines and possibly in animal tumor models. 1: Kanwal, U., et. al. (2018). Advances in nano-delivery systems for doxorubicin: an updated insight. Journal of drug targeting, 26(4), 296-310 2: Barenholz, Y. C. (2012). Doxil®�the first FDA-approved nano-drug: lessons learned. Journal of controlled release, 160(2), 117-134. 3: Matuszcak, C., et. al. (2015). Proton pump inhibitors as chemosensitizer: New indication for a well-known medication. Cancer Cell & Microenvironment, 2(1). 4: Yu, M., et. al. (2015). Influence of the proton pump inhibitor lansoprazole on distribution and activity of doxorubicin in solid tumors. Cancer science, 106(10), 1438-1447.


11

11

A MICROSERVICE-BASED INFRASTRUCTURE FOR LIGAND-BASED MODELING

Jonathan Alvarsson (a), B Ernst Ahlberg Helgee (b), A. Arvid Berg (a), A. Ola Spjuth (a) [email protected]

(a) Department of Pharmaceutical Biosciences, Uppsala University, (b) Predictive Compound ADME &

Safety, Drug Safety & Metabolism, AstraZeneca IMED Biotech Unit

New data is continously being generated and prediction models should be rebuilt to keep up. We are building a microservice-based system for producing and publishing ligand-based prediction models. Models are built in a workflow that can be re-run when data is updated. The model is then packaged and published as a microservice in a Docker container in Kubernetes / OpenShift. The model can be accessed in a web based user interface or used in scripts via a REST based API published according to the Swagger standard.


12

12

A HIGH-THROUGHPUT METHOD TO PREDICT THE IN VIVO EXPOSURE OF DIFFERENT DRUG DELIVERY SYSTEMS

C. Alvebratt (a), J. Keemink (a), K. Edueng (a), O. Cheung (b), M. Strømme (b), C. A.S. Bergström (a)

[email protected]

(a) Department of Pharmacy, Uppsala University, Sweden, (b) Department of Engineering Sciences, Uppsala University, Sweden

Supersaturating drug delivery systems (SDDS) have been suggested as a promising strategy to overcome poor aqueous solubility and hence increase in vivo exposure of drug compounds. Commonly used SDDS include amorphous solid dispersions (ASD), carrier-based delivery systems and lipid-based formulations (LBF). Methods used to determine release and absorption from these delivery systems are unfortunately significantly different, which limits the possibilities to perform head-to-head comparisons of formulations. In this work, we present a method that allows real time monitoring of release and absorption of felodipine from several different SDDS. The µDiss ProfilerTM was combined with a slightly modified µFLUXTM system, which herein was based on Caco-2 cell monolayers as the absorptive membrane. To reduce particle interference with analytics, which in this system is based on in situ UV-probes, a protective filter composed on nylon was placed on the UV-probe in the donor chamber [1]. The release and absorption were determined from crystalline drug, three types of LBFs, one ASD and one drug-loaded mesoporous carrier. A modified lipolysis buffer was used to ensure that the pH would remain constant (pH 6.5) during digestion of the LBFs [2]. One of the LBFs showed extensive absorption at the same wavelength as felodipine, and therefore HPLC-UV was utilized to determine the donor concentrations of this LBF. Significant differences with regard to degree of supersaturation reached in the donor compartments were observed between the SDDSs. The absorption rate was despite that similar for all formulations; the supersaturation rank order was not directly related to the flux across the membrane. To conclude, a method that enables monitoring and comparison of the release and permeation behavior of felodipine from various SDDs has been developed. The method is a valuable addition to currently available dissolution and absorption models, and is applicable to a wide range of drug delivery systems. References: [1] Alvebratt et al., 2018, AAPS PharmSciTech. 1-7 [2] Mosgaard et al., 2015, Eur J Pharm Biopharm, 94:493-500.


13

13

SYNERGY CONFORMAL PREDICTION

Niharika Gauraha, Ola Spjuth [email protected]

Department of Pharmaceutical Biosciences, Uppsala University

Conformal Prediction is a machine learning methodology that produces valid prediction regions under mild conditions. Ensembles of conformal predictors have been proposed to improve the informational efficiency of inductive conformal predictors by combining p-values, however, the validity of such methods has been an open problem. We introduce Synergy Conformal Prediction, which is an ensemble method that combines monotonic conformity scores, and is capable of producing valid prediction intervals. We study the applicability in two scenarios; where data is partitioned in order to reduce the total model training time, and where an ensemble of different machine learning methods is used to improve the overall efficiency of predictions. We evaluate the method on 10 data sets and show that the synergy conformal predictor produces valid predictions and improves informational efficiency as compared to inductive conformal prediction and existing ensemble methods. The results indicate that synergy conformal prediction has advantageous properties compared to contemporary approaches, and we also envision that it will have an impact in Big Data and federated environments.


14

14

ISOLATION AND IDENTIFICATION OF THE MAIN STEROL FROM THE SPONGE HALICLONA SP. FROM TAIWAN

Merran Dunford (a,b), Sanjeevan Ranjendran (a), Patric Jonsson (a), Ping-Jyun Sung (c), Fang-Rong

Chang (d), Achyut Adhikari (e), Sunithi Gunasekera (a), Paco Cárdenas (a) [email protected]

(a) Pharmacognosy, Department of Medicinal Chemistry, BMC, Uppsala University, Uppsala, Sweden, b)

Department of Pharmacy and Pharmacology, University of Bath, Bath BA2 7AY, UK, c) National Museum of Marine Biology and Aquarium, Pingtung 944, Taiwan, d) Graduate Institute of Natural

Products, College of Pharmacy, Kaohsiung Medical University, Kaohsiung 80708, Taiwan, e) International Centre for Chemical and Biological Sciences, H. E. J. Research Institute of Chemistry, University of

Karachi, Pakistan

Sponges are known to harbor a wide spectrum of bioactive compounds of potential pharmaceutical value. In the current study, sterols were isolated from the shallow water Haliclona sp. from South Taiwan, and their antimicrobial activity was investigated. A normal phase, open-column chromatography procedure was adopted to isolate sterols from the sponge extract. From the initial NMR experiments, the most abundant compound in the sponge extract was suggested to be ß-sitosterol present in a mixture with several other sterols. Subsequently a preparative TLC was utilized to further separate the sterol mixture by using ß-sitosterol as a reference. The NMR analysis of the isolated peak having similar Rf value as ß-sitosterol from the preparative TLC procedure, gave evidence of a partially purified sterol mixture with the presence of a sterol with an additional double bond to ß-sitosterol. The partially purified sterol mixture was determined to have no activity on Escherichia coli and Staphylococcus aureus at a concentration of less than 0.5 mg/mL. Financial support from the EU Erasmus programme, the EU Horizon 2020 programme SponGES under grant agreement No. 679849 and an International Collaborative Research Grant (Swedish Research Council).


15

15

A DIGESTION/PERMEATION MODEL BASED ON ARTIFICIAL MEMBRANES FOR STUDIES OF PERFORMANCE OF LIPID-BASED FORMULATIONS

O. J. Hedge, C. A. S. Bergström

[email protected]

Department of Pharmacy, Uppsala University, Uppsala Biomedical Center, P.O. Box 580, SE-751 23 Uppsala, Sweden

Lipid-based formulations (LBFs) may be a viable strategy for oral delivery of poorly water-soluble drug compounds, since they may increase bioavailability and reduce food effects of many contemporary drug molecules. There is, however, a lack of in vitro performance assessment assays that correlate well to in vivo data. One reason could be focus on precipitation during dispersion and digestion in simulated intestinal fluid (SIF), without taking absorption into account (1). Attempts to incorporate simultaneous absorption into the lipolysis assay have been made using cell monolayers (2). While these are promising, the cells are sensitive to the enzymes used and to date, digestion in the presence of cells is based on immobilized recombinant lipase that is not reflecting the diversity of the pancreatic enzymes. We therefore identified an artificial membrane that withstands the harsh environment of LBF digestion. Polymer filter supports were treated with oil/lipid solutions to form artificial membranes. LBFs were dispersed in fasted state SIF, spiked with the impermeable marker Lucifer Yellow (LY), and sampled during dispersion and digestion. Digestion was initiated after 10 minutes of dispersion, through addition of porcine pancreatic extract. Formulations of three different LBF classes loaded with 80 mg/g fenofibrate (FEN) were assayed using the novel ENA system with sampling at defined time points from donor and receiver compartments. Donor samples were filtered (0.1 µm) to separate precipitates from the matrix. Samples were quantified for FEN using LC-MS and LY using fluorescence detection. Of the physical membranes explored, LIDO applied to PVDF membranes gave the lowest flux of LY between the ENA compartments. The model produced significant (p < 0.05) differences in flux of FEN between the formulations. The rank order of the flux between the formulations was comparable to that of the fraction FEN solubilized in the donor compartment. While the system could withstand the digestion process of several types of LBFs, the in vitro in vivo correlation (IVIVC) was not improved as compared to the previously explored methods (3). However, previous studies in the ENA system utilizing Caco-2 monolayers have shown rank order IVIVC for the formulations tested here (4). As FEN is not a known transporter substrate, this could indicate other mechanisms at play, e.g. uptake of lipids and bile salts. Further studies comparing the structural evolution of the donor matrix over time in a cell-based model could therefore be required to elucidate the mechanics of LBF mediated drug absorption. References: (1) O. M. Feeney et al., Adv. Drug Deliv. Rev. 101, 167�194 (2016). (2) J. Keemink, C. A. S. Bergström, Pharm. Res. 35 (2018), doi:10.1007/s11095-017-2327-8. (3) B. T. Griffin et al., Eur. J. Pharm. Biopharm. 86, 427�437 (2014). (4) J. Keemink et al., Submitted manuscript (2018).


16

16

MAPPING THE DISTRIBUTION OF NEUROPEPTIDES THROUGH MASS SPECTROMETRY IMAGING

Heather E Hulme (a), Halla Gunnarsdóttir (a), Elva Fridjonsdottir (a), Anna Nilsson (a),

Mohammadreza Shariatgorji (a), Erwan Bezard (b), Per E. Andrén (a) [email protected]

(a) Medical Mass Spectrometry Imaging, National Resource for Mass Spectrometry Imaging, Science for

Life Laboratory, Department of Pharmaceutical Biosciences, Uppsala University, Box 591 BMC, 75124 Uppsala, (b) Université de Bordeaux, Institut des Maladies Neurodégénératives,

UMR 5293 Bordeaux, France

Neuropeptides are a group of signalling molecules, which are produced by neurons and regulate neural function. Neuropeptides are particularly interesting as changes in the levels of these molecules has been associated with various neurological disorders such as Parkinson’s disease. Matrix assisted laser desorption ionisation mass spectrometry imaging (MALDI-MSI) is a beneficial technique to investigate neuropeptides as this technique can map the distribution of various molecules at once across a thin tissue sample. Therefore, this technique has the potential to image various neuropeptides at once across the brain to help further understand neuropeptides and their role in neurological disorders. We optimised a method to enhance the detection of neuropeptides by MALDI-MSI, which allowed the imaging of various neuropeptides that have not been previously visualised by MSI. This method was subsequently applied to map the distribution of neuropeptides in rodent and primate brains as well as other tissue types. MALDI-MSI was further used to compare the neuropeptide abundance and distribution in experimental models of Parkinson’s disease compared to healthy controls. This study highlights the capabilities of MALDI-MSI to help further understand neuropeptides and their role in neurological disorders.


17

17

DOCKING OF MACROCYCLES: COMPARING RIGID AND FLEXIBLE DOCKING IN GLIDE

Hiba Alogheli, Gustav Olanders, Wesley Schaal, Peter Brandt, Anders Karlén

[email protected]

Uppsala university

In recent years, there has been an increased interest in using macrocyclic compounds for drug discovery and development. For docking of these commonly large and flexible compounds to be addressed, a screening and a validation set were assembled from the PDB consisting of 16 and 31 macrocycle-containing protein complexes, respectively. The macrocycles were docked in Glide by rigid docking of pregenerated conformational ensembles produced by the macrocycle conformational sampling method (MCS) in Schrödinger Release 2015-3 or by direct Glide flexible docking after performing ring-templating. The two protocols were compared to rigid docking of pregenerated conformational ensembles produced by an exhaustive Monte Carlo multiple minimum (MCMM) conformational search and a shorter MCMM conformational search (MCMM-short). The docking accuracy was evaluated and expressed as the RMSD between the heavy atoms of the ligand as found in the X-ray structure after refinement and the poses obtained by the docking protocols. The median RMSD values for top-scored poses of the screening set were 0.83, 0.80, 0.88, and 0.58 Å for MCMM, MCMM-short, MCS, and Glide flexible docking, respectively. There was no statistically significant difference in the performance between rigid docking of pregenerated conformations produced by the MCS and direct docking using Glide flexible docking. However, the flexible docking protocol was 2-times faster in docking the screening set compared to that of the MCS protocol. In a final study, the new Prime-MCS method was evaluated in Schrödinger Release 2016-3. This method is faster compared that of to MCS; however, the conformations generated were found to be suboptimal for rigid docking. Therefore, on the basis of timing, accuracy, and ease of set up, standard Glide flexible docking with prior ring-templating is recommended over current gold standard protocols using rigid docking of pregenerated conformational ensembles.


18

18

HOME VS. SELF-INITIATED ART REFILL: CLINICAL, IMMUNOLOGICAL, AND VIROLOGIC OUTCOMES

Rory F. Leisegang (1,2), Jane Ball (3), Mark Cotton (4), David Dowdy (5), Keri Calkins (5), Susan

Cleary (2), Gary Maartens (2), Jean B. Nachega (6) [email protected]

(1) Uppsala University, Uppsala; (2) University of Cape Town, Cape Town, South Africa, (3) AID for

AIDS, Cape Town, South Africa; (4) Stellenbosch University, Tygerberg, South Africa; (5) Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA; (6) University of Pittsburgh, Pittsburgh, PA,

USA

Background: Antiretroviral therapy (ART) delivery by courier to the patient's home (home refill) is a novel intervention that may improve clinical outcomes and reduce indirect costs for individuals in low- and middle-income countries (LMICs). We aimed to compare clinical and virologic outcomes for patients obtaining medication refills at their local pharmacy (self refill) vs. home refill in Aid for AIDS (AFA), a large South African private sector HIV/AIDS programme. Methods: Retrospective cohort analysis of ART naïve HIV-infected adults in AFA who initiated first line NNRTI based ART regimen between January 2002 and July 2010 was performed. Patients were selected to switch to home refill based on the discretion of AFA. Primary endpoint was all-cause mortality; secondary endpoints were viral suppression (VL< 400 copies/mL) and median CD4+ T-cell response (cells/µl) (from baseline) at 6-month intervals. We compared the crude survival between self-refill and home refill using Kaplan�Meier plots and a log-rank test. We performed Cox regressions to model the individual and simultaneous effects of baseline variables and mode of ART delivery on all-cause mortality, adjusting for bias using propensity score and marginal structural models. Results: 40,939 patients, contributing 66,204 years of follow-up were recorded. The most common first line regimen was efavirenz + lamuvidine + zidovudine, followed by efavirenz + emtricitabine + tenofovir in later years. Emerging at 24 months, the home refill group had improved median CD4+ T-cell count response (451 vs. 387, respectively, p < 0.01), and the likelihood of virologic suppression (81% versus 71%, respectively, p-value <0.001), compared to the self-refill group. Home refill (vs. self-refill) was associated with better survival (adjusted hazard ratio = 0.90 [95% CI: 0.84-0.96], p-value for log-rank test < 0.001) (Figure 1). Conclusion: Home refill is associated with improved clinical, immunological, and virologic outcomes compared to self-refill for HIV-infected adults in this private AIDS programme in South Africa. Home refill offers a promising additional option to the growing ART service delivery models and should facilitate the UNAIDS 90-90-90 targets in LMICs.


19

19

CHEMOSELECTIVE PROBES FOR THE DETECTION OF MICROBIAL BIOMARKERS OF DISEASE - COMBINING METABOLOMICS WITH CHEMICAL BIOLOGY

METHODOLOGIES

Neeraj Garg, Louis P. Conway, Caroline Ballet, Mário S. P. Correia, Daniel Globisch. [email protected]

Department of Medicinal Chemistry, Uppsala University

The discovery of highly specific small-molecule biomarkers of disease is the first step in the development of rapid non-invasive diagnostics, permitting the routine screening of at-risk populations and detection of disease at an early, treatable stage. In recent years, the gut microbiota has been found to profoundly impact human health. While the co-metabolism between commensal bacteria and the human host is potentially a rich source of biomarkers of disease, this chemical environment remains relatively unexplored. LC-MS-based general metabolomics is well suited to the sensitive detection of the vast number of compounds present in a biological matrix, however, the structural identification of these metabolites has long presented a major bottleneck. Furthermore, metabolites which occur at low concentrations are often neglected entirely due to compromises in analytical methodology or the lack of appropriate chemical tools. We have therefore developed chemical biology-based probes for the highly selective modification of well-defined metabolite classes, enhancing their sensitivity under LC-MC analysis and facilitating their identification. We seek to apply these tools to the analysis of patient-derived samples, with a particular focus on disease-microbiome interactions and cancer. This study was supported by the Swedish Research Council (VR 2016-04423), the Carl Tryggers Foundation (CTS16:155),and the Science for Life Laboratory to D.G


20

20

LCMS-BASED UNTARGETED METABOLOMICS USING GUINEA PIG PERILYMPH: HIGHLIGHTING DIFFERENCES IN METABOLOME WITH AND WITHOUT A

POSTERIORI HYDROGEN GAS ADMINISTRATION FOLLOWING LOUD NOISE EXPOSURE

K. Pirttilä (a), A. E. Fransson (b), J. Haglöf (a), M. Engskog (a), P. V. Pierre (c), G. Laurell (b),

T. Arvidsson (a, d), C. Pettersson (a) [email protected]

(a) Dept. Medicinal Chemistry, Uppsala University, Uppsala, Sweden. (b) Dept. of Surgical Sciences, Uppsala University, Uppsala, Sweden. (c) Div. of Audiology, Dept. Clinical science, Intervention, and Technology, Karolinska Institutet, Stockholm, Sweden. (d) Medical Product Agency, Uppsala, Sweden

Introduction. In metabolomics studies of inherently localized conditions, the use of systemic matrices such as blood plasma risks diluting metabolic differences, possibly rendering them undetectable. In such studies, the use of localized tissue, sampled in proximity to the afflicted region is preferable. In the study of conditions of the inner ear, such as noise-induced hearing loss (NIHL), the use of perilymph aspired directly from the inner ear allows a more veracious view of the current metabolic state of the studied system. In this study a protocol is presented for untargeted metabolomics of perilymph aspired from the scala tympani of guinea pigs. The protocol is further applied to study the protective effect of gaseous hydrogen on NIHL in guinea pig. Methods. Animals were exposed to noise (Noise, n=8) (156 dB, 400 impulses over 3 min) or noise followed by H2 gas administration (Noise+H2, n=9). Perilymph (1 µL) was aspired from the basal turn of the noise-exposed cochlea and diluted 1:19 with water. Prior to LCMS analysis the perilymph was subjected to protein precipitation. The supernatant was injected without further treatment. UPLC-Q-TOF (Waters, Milford, USA) analysis was achieved using a BEH Amide column (50x2.1mm, 1.7µm, Waters) operated in HILIC mode. Raw data was subjected to pretreatment by peak picking (XCMS), normalization by median fold change, and filtering of features with high variability in repeat injections. Multivariate data analysis was performed on pareto scaled data in the SIMCA software (Umetrics, Umeå, Sweden). Preliminary data. Following normalized and filtering, a total of 689 features were retained in using positive ionization mode corresponding to 276 peak groups as determined by CAMERA. Relative standard deviation (RSD) of peak areas for 8 known stable metabolites ranged between 3.4-12.3% across all QC injections in the analytical run (n=15). Retention time RSD of the same metabolites ranged from 0.00-0.34%. Multivariate data analysis of the resulting dataset allowed a clear discrimination between the sample groups Noise and Noise+H2, whereas the Noise+H2 group appears more similar to the control groups Control (n=2, no treatment), and H2 (n=3, only hydrogen gas treatment). Principal component analysis (PCA) shows the Noise group as a distinct cluster that deviates from the other sample groups. Supervised analysis by orthogonal partial least squares discriminant analysis (OPLS-DA) on the sample groups Noise and Noise+H2 shows excellent model performance indicators (R2 0.968, Q2 809) as expected from the PCA model. The indicated features were further considered manually to filter out adducts and non-peak noise data. Univariate analysis by one-sided Wilcoxon rank-sum tests of the selected 15 metabolites showed that many of them were significantly different in the two noise exposed groups. A common pattern was also that the Noise+H2 group had similar metabolite levels as the two control sample groups, indicating that the hydrogen gas treatment has a preventive effect on the metabolic dysregulation caused by the noise exposure.


21

21

POPULATION MODELING OF UNI-AND THREE-DIMENSIONAL AND DENSITY-BASED TUMOR MEASUREMENTS IN GASTRO-INTESTINAL STROMAL TUMOR

PATIENTS TREATED WITH IMATINIB

Sreenath M. Krishnan (a), Emilie Schindler (a), Gaia Schiavon (b,c), Ron Mathijssen (b), Lena E. Friberg (a)

[email protected]

(a) Department of Pharmaceutical Biosciences, Uppsala University, Sweden. (b) Department of Medical Oncology, Erasmus University Medical Center, Rotterdam, The Netherlands.

(c) Translational Science, Oncology iMed, AstraZeneca, Cambridge, UK. Objectives: Three-dimensional (3D) and density-based tumor metrics have been suggested to better discriminate tumor response to treatment than the traditional unidimensional (1D) metrics for gastro-intestinal stromal tumor (GIST) which often exhibit non-uniform size changes. This study aims to characterize the longitudinal 1D, 3D and density responses of liver metastases in imatinib-treated GIST patients and quantify the inter-individual (IIV) and inter-lesion variability (ILV) and investigate model-predicted tumor metrics as predictors of overall survival (OS) and time to treatment failure (TTF). Methods: Data were obtained from a retrospective, non-interventional study involving 77 GIST patients treated with oral imatinib at a starting dose of 400 mg/800 mg q.d. (74/3) [1]. Maximum trans-axial diameter (MTD), software-calculated segmented volume (Vs), calculated ellipsoidal volume (Ve) and density data were collected from 137 lesions (1-2 per patient) at baseline and at least one post-baseline visit with the median follow-up time of 360 days. Tumor growth inhibition (TGI) models with constant, exponential or Gompertz growth were explored to describe MTD, Vs and Ve data. Indirect response models (IDR) with inhibition of the production or stimulation of the loss of response were investigated for tumor density. IIV and ILV were tested in all parameters. All models were combined into a joint model to explore correlations. Parametric time-to-event (TTE) models were developed for exploring potential predictors for OS and TTF. Results: Tumor modeling: TGI models with an exponential growth and a linear dose-driven drug effect that washes out over time [2] best characterized the MTD, Vs and Ve data. A mixture model described the bimodal distributions of the baseline data and accounted for ILV. The model-predicted doubling time was typically lower for Ve (1.3 years) and Vs (1.5 years) than for MTD (8.7 years). The growth rate constants and drug effect parameters were associated with large IIV but no ILV. An IDR model with stimulation of the loss of response adequately described tumor density data. Large IIV (120 %CV) and ILV (53 %CV) were identified in the drug effect. High correlations (>99%) were estimated between MTD, Vs and Ve model parameters. Correlations between the density model parameters and MTD, Vs and Ve model parameters were low (<20%). OS and TTF analysis: In the multivariate analysis, statistically significant predictors (p<0.01) of OS included baseline MTD (hazard ratio (HR) =1.24 for every 10 mm increase) and relative MTD change from baseline at 6months (HR=1.03 for 10% increase from baseline). Statistically significant predictors (p<0.01) of TTF included baseline MTD (HR=1.16 for every 10 mm increase) and relative Ve change from baseline at 3months (HR=1.03 for 10% increase from baseline). Conclusions: The developed models adequately described the longitudinal 1D, 3D and density data of liver metastases in imatinib-treated GIST patients. All tumor responses were associated with large IIV. ILV was identified in the drug sensitivity of density only. Smaller MTD at baseline were associated with longer OS and TTF. Moreover, a larger decrease in MTD at 6 months was predictive of a longer OS and a larger decrease in Ve at 3 months of a longer TTF. Acknowledgements: This work was supported by the Swedish Cancer Society. References: [1] Schiavon G. et al. Eur J Cancer (2014) 50:972-80. [2] Claret et al. J Clin Oncol (2009) 27: 4103-4108.


22

22

POPULATION PHARMACOKINETIC ANALYSIS OF FACTOR VIII ACTIVITY FOLLOWING TREATMENT WITH MOROCTOCOG ALFA IN MODERATE TO

SEVERE HAEMOPHILIA A SUBJECTS

João A. Abrantes (a), Elisabet I. Nielsen (a), Joan Korth-Bradley (b), Lutz Harnisch (c), Siv Jönsson (a) [email protected]

(a) Department of Pharmaceutical Biosciences, Uppsala University, Uppsala, Sweden, (b) Pfizer Inc,

Collegeville, Pennsylvania, USA, (c) Global Clinical Pharmacology, Pfizer, Sandwich, UK

Objectives: Moroctocog alfa is a B-domain deleted recombinant factor VIII (FVIII) indicated for the treatment and prophylaxis of bleeding in adults and children with haemophilia A. The aim of this study is to develop a population pharmacokinetic model of FVIII activity following moroctocog alfa treatment in moderate to severe haemophilia A patients using all available clinical study data. Methods: A population pharmacokinetic model (NONMEM 7.3) was developed with data from 13 trials involving IV administration of moroctocog alfa products (Refacto®, Refacto AF®, Xyntha® Wyeth Pharmaceuticals Inc. [Pfizer], Philadelphia, USA). The studies were conducted from 1993 to 2013 in 25

countries, and included rich sampling (≥10 samples post-dose, n=4), sparse sampling (2-3 samples per occasion, n=5) or both (n=4). The influence of age, measures of body size, race, ethnicity, inhibitor status and titre, assay, year of study, study and country, were investigated. Data below the lower limit of quantification were handled using the M5 method [1] and correlated residual errors for repeated analyses were accounted for by using the L2 data item [2]. Qualification of the models included pcVPCs and parameter uncertainty was obtained with SIR [3]. Results: A total of 259 children and adolescents and 497 adults with moderate to severe haemophilia A

(FVIII ≤5 IU/dL) were included in the analysis. Age ranged from 1 day to 73 years and weight from 3 to 134 kg. A two-compartment model with allometrically scaled body weight on disposition parameters was found to adequately describe the combined data. Pre-dose activity observations were described by estimating two endogenous FVIII activity levels at 0.474 IU/dL and 1.59 IU/dL together with an additive component accounting for residual activity from previous unknown doses (2.84 IU/dL at time 0, declining over time in line with FVIII disposition). For a 70-kg, 20-year-old patient with moderate or severe haemophilia A, FVIII activity clearance was estimated at 2.76 dL/h, central volume of distribution at 24.5 dL, peripheral volume of distribution at 9.23 dL, and inter-compartmental clearance at 25.1 dL/h. Conclusions: The model is capable of describing and predicting reasonably well the data and when used with patient observations in a therapeutic drug monitoring context may be an aid in dose individualization, with the ultimate goal of a safer and more effective treatment with moroctocog alfa. References: [1] Ahn JE, Karlsson MO, Dunne A, Ludden TM. 2008. Likelihood based approaches to handling data below the quantification limit using NONMEM VI. J Pharmacokinet Pharmacodyn 35(4):401-21 [2] Karlsson MO, Beal SL, Sheiner LB. 1995. Three new residual error models for population PK/PD analyses. J Pharmacokinet Biopharm 23:651�672 [3] Dosne AG, Bergstrand M, Karlsson, MO. Application of Sampling Importance Resampling to estimate parameter uncertainty distributions. PAGE 22 (2013) Abstr 2907 [www.page-meeting.org/?abstract=2907]


23

23

PROBABILISTIC PREDICTIONS OF METABOLISM USING VENN-PREDICTORS

Staffan Arvidsson Mc Shane (a), Lars Carlsson (b), Paolo Toccaceli (c), Ola Spjuth (a) [email protected]

(a) Department of Pharmaceutical Biosciences, Uppsala University, (b) Quantitative Biology, Discovery

Sciences, Innovative Medicines & Early Development, AstraZeneca, (c) Department of Computer Science, Royal Holloway

Prediction of drug metabolism is an important topic in the drug discovery process, and we here present a study using probabilistic predictions applying Cross Venn-ABERS Predictors (CVAPs) on data for site-of-metabolism. We used a dataset of 73599 biotransformations, applied SMIRKS to define biotransformations of interest and constructed five datasets where chemical structures were represented using signatures descriptors. The results show that CVAP produces well-calibrated predictions for all datasets with good predictive capability, making CVAP an interesting method for further exploration in drug discovery applications.


24

24

CONTRALESIONAL HINDLIMB MOTOR RESPONSE INDUCED BY UNILATERAL BRAIN INJURY: EVIDENCE FOR EXTRA- SPINAL MECHANISM

N. Lukoyanov (a), L. Carvalho (a), H. Watanabe (b), M. Zhang (c, d), D. Sarkisyan (b), O. Kononenko

(b), I. Bazov (b), T. Iakovleva (b), J. Schouenborg (c) and G. Bakalkin (b) [email protected]

(a) Instituto de Investigação e Inovação em Saúde, Instituto de Biologia Molecular e Celular,

Departamento de Biomedicina da Faculdade de Medicina da Universidade do Porto, Portugal; (b) Dept. Pharm. Biosciences, Uppsala University, Sweden; (c) Neuro-nano Research Center, Lund University,

Sweden; and (d) Dept. Mol. Medicine, University of Southern Denmark, Odense, Denmark

INTRODUCTION. The neurological dogma states that motor deficits secondary to traumatic brain injury and stroke arise due to the aberrant activity of neural pathways descending from injured brain to the spinal cord. OBJECTIVE. We tested this dogma by analyzing hindlimb motor response to unilateral brain injury in the rats with completely transected spinal cord. METHODS. The spinal cord was transected at T2/T3 levels first, and then the cortical hindlimb representation area was unilaterally ablated. Motor outputs were analyzed under pentobarbital anesthesia during 3-4 hours. Formation of hindlimb postural asymmetry was assessed as differences between the contra- and ipsilesional legs in a) their position; b) hip, knee and ankle joint angles, and (c) stretch forces. The nociceptive withdrawal reflexes, quantified as EMG responses evoked by electrical stimulation of digits and heel, were compared between the ipsi- and contralesional hindlimb muscles. Expression of plastic genes was analyzed in the lumbar spinal cord, i.e. below the transection level by qRT-PCR. RESULTS. Behavioral, electrophysiological and molecular evidence demonstrated the development of the contralesional-side specific response to the unilateral brain injury. The spinal rats, which then received cortical ablation but not sham operation, developed asymmetric hindlimb posture with flexion of the contralesional leg. The hindlimb withdrawal reflexes of the flexor muscles (the extensor digitorum longus and semitendinosus) were enhanced on the contralesional side while extensor muscle (the interosseous) on the ipsilesional side. Unilateral cortical ablation in spinal rats produced robust changes in the expression of neuroplastic genes in the lumbar spinal cord. mRNA levels were elevated on either the ipsilesional (c-Fos) or contralesional (TGF-beta) side resulting in the asymmetric expression patterns, or bilaterally (Gap43). Unilateral brain injury failed to induce postural asymmetry in the hypophysectomized spinal rats. CONCLUSIONS. These results demonstrate that, in parallel with neural pathways, signals elicited by injured brain are transmitted to the lumbar spinal cord by alternative, likely endocrine mechanism. Spinal plastic and hindlimb motor responses induced by these signals are side specific.


25

25

EXPLORING THE USEFULNESS OF MORPHOLOGICAL PROFILING OF CELLS TO STUDY TOXICITY MECHANISMS

P. Georgieva, Tanya Aggarwal, W. Schaal, O. Spjuth

[email protected]

Uppsala University, Department of Pharmaceutical Biosciences, Uppsala, Sweden Although there have been major technological advances in how we study biological processes and systems in the last years, studying toxicity pathways still remains difficult, time-consuming and resource-intensive. High-content imaging of cells stained with multiplexed dyes, “Cell Painting”(1), has been successfully applied in drug screening(2) of chemical compounds based on morphological changes they induce. However its application in the field of toxicity is still rather unexplored and comprises an exciting opportunity to unravel toxicity mechanisms and pathways. We are investigating the usefulness of image-based morphology profiling toward understanding toxicology mechanisms and pathways, by studying a range of cell lines treated with different toxicants. To this end we implemented a cell painting screening assay, involving 4 - 6 fluorescent dyes, used to stain at least 6 - 8 subcellular components simultaneously, followed by subsequent imaging with an ImageXpress XLS high-content microscope. To analyse these compound-induced morphological changes we use a “CellProfiler”(3) analysis pipeline to compare morphological profiles and classify compounds based on mechanisms. Our initial experiments show promising results for mechanistic classification of toxicity mechanisms and we hypothesize that the method could be used for pathway enrichment. To this end we will strive towards automating the platform, and develop robust analysis pipelines to handle data/analysis to be able to study mechanisms on larger scale (1) Bray, Mark-Anthony et al. "Cell Painting, A High-Content Image-Based Assay For Morphological Profiling Using Multiplexed Fluorescent Dyes". Nature Protocols 11.9 (2016): 1757-1774. (2) Simm, Jaak et al. “Repurposing High-Throughput Image Assays Enables Biological Activity Prediction for Drug Discovery”. Cell Chemical Biology, 25.5 (2018), 611-618 (3) Carpenter, Anne et al. “CellProfiler: image analysis software for identifying and quantifying cell phenotypes”. Genome Biology (2006) 7:R100


26

26

EXPLORING THE USEFULNESS OF MORPHOLOGICAL PROFILING OF CELLS TO STUDY TOXICITY MECHANISMS

P. Georgieva, T. Aggarwal, W. Schaal, O. Spjuth

[email protected]

Uppsala University, Department of Pharmaceutical Biosciences, Uppsala, Sweden Although there have been major technological advances in how we study biological processes and systems in the last years, studying toxicity pathways still remains difficult, time-consuming and resource-intensive. High-content imaging of cells stained with multiplexed dyes, “Cell Painting”(1), has been successfully applied in drug screening(2) of chemical compounds based on morphological changes they induce. However its application in the field of toxicity is still rather unexplored and comprises an exciting opportunity to unravel toxicity mechanisms and pathways. We are investigating the usefulness of image-based morphology profiling toward understanding toxicology mechanisms and pathways, by studying a range of cell lines treated with different toxicants. To this end we implemented a cell painting screening assay, involving 4 - 6 fluorescent dyes, used to stain at least 6 - 8 subcellular components simultaneously, followed by subsequent imaging with an ImageXpress XLS high-content microscope. To analyse these compound-induced morphological changes we use a “CellProfiler”(3) analysis pipeline to compare morphological profiles and classify compounds based on mechanisms. Our initial experiments show promising results for mechanistic classification of toxicity mechanisms and we hypothesize that the method could be used for pathway enrichment. To this end we will strive towards automating the platform, and develop robust analysis pipelines to handle data/analysis to be able to study mechanisms on larger scale (1) Bray, Mark-Anthony et al. "Cell Painting, A High-Content Image-Based Assay For Morphological Profiling Using Multiplexed Fluorescent Dyes". Nature Protocols 11.9 (2016): 1757-1774. (2) Simm, Jaak et al. “Repurposing High-Throughput Image Assays Enables Biological Activity Prediction for Drug Discovery”. Cell Chemical Biology, 25.5 (2018), 611-618 (3) Carpenter, Anne et al. “CellProfiler: image analysis software for identifying and quantifying cell phenotypes”. Genome Biology (2006) 7:R100


27

27

DISCREPANCY IN STABLY EXPRESSED PROTEINS WITHIN AND ACROSS HUMAN TISSUE TYPES

C. Wegler (a,b), M. Ölander (a), P. Artursson (a)

[email protected]

(a) Department of Pharmacy, Uppsala University, Uppsala, Sweden (b) Drug Metabolism and Pharmacokinetics, Cardiovascular, Renal and Metabolism, IMED Biotech Unit,

AstraZeneca Gothenburg, Gothenburg, Sweden In recent years, large-scale omics analyses have mapped human gene and protein expression to characterize and compare different cells and tissues. It has been shown that some proteins are mostly expressed in a limited number of tissues, while others are widely and uniformly distributed throughout the human body. Such uniform proteins have been proposed as references for quality control and normalization of omics data, as well as loading controls for blotting techniques. However, these studies did not specifically consider within-tissue variability across human subjects and its impact on normalization in the analysis of the same tissue type from many individuals. Here, we therefore analyze inter-sample variability in proteomic datasets from human liver. We describe the characteristics of the least and most variable proteins, and provide a set of proteins with uniform expression across tissue samples. Surprisingly, many previously described proteins with low variability across tissue types showed large within-tissue variability. We observed a wide distribution in variabilities, with highly abundant proteins showing less variable expression across individuals. The least variable proteins within tissues were generally confined to the cytosol or mitochondria, while the most variable proteins were more spread across different subcellular compartments. Further, proteins essential for cell survival were overrepresented in the low variability range, suggesting that survival requires expression of these proteins at a certain level in all individuals. Supporting this, inter-sample variability showed an inverse correlation with previously established protein turnover rates. In conclusion, the differences in variability within and across tissues highlight the difficulties in selecting single proteins for use as references in all tissue types.


28

28

EARLY-LIFE EXPOSURE TO TO AN ALGAL NEUROTOXIN INDUCES LONG-TERM CEREBELLAR NEURODEGENERATION AND LYSOSOMAL CHANGES

J Swan (a), L Ersson(a), A-L Berg(b), O Karlsson(a,c), E Brittebo(a)

[email protected]

(a) Dept Pharm. Biosciences, Uppsala University, (b) Läkemedelsverket, (c) Stockholm University

Algal blooming is of increasing concern with respect to cyanobacteria (blue-green algae) since it is promoted by eutrophication of aquatic environments and global warming. Most cyanobacteria can produce the neurotoxin ß-N-methylamino-L-alanine (BMAA). Dietary exposure to this neurotoxin has been suggested to be involved in the etiology of sporadic neurodegenerative disease. Current research on the etiology of neurodegenerative disease is mainly focused on identifying gene mutations in the familiar forms of disease. The cause for the majority of sporadic cases (also called late onset) of neurodegenerative disease is still unknown. There is, however, considerable evidence to support a multifactorial etiology involving both environmental, life-style and genetic factors. Our previous studies revealed high transfer of BMAA to the neonatal rodent brain with a distinct localization in the hippocampus and cerebellum. Early-life exposure to the neurotoxin BMAA during the brain development period may trigger cell stress and perturbations of cell signaling and protein expression in sensitive brain regions, which eventually translates into progressive neurodegeneration at adult age. Our studies on the effects of BMAA in the hippocampus demonstrated that early-life exposure to BMAA induced behavioual changes such as deficits in learning and memory functions as well as a selective long-term neurodegeneration, calcification and intracellular fibril formation in the CA1 region. We are now studying the long-term effects of BMAA in the cerebellum following a short exposure during the neonatal period. The results demonstrated a selective long-term neurodegeneration, calcification as well as changed protein expression in the cerebellar vermis region. Studies on perturbation of protein expression in this area suggest lyzosomal changes.


29

29

MASS SPECTROMETRY LIPIDOMICS: A PLATFORM TO UNRAVEL LIPID BIOLOGY

David Balgoma (a), Alfhild Grönbladh (b), Sofia Zelleroth (b), Mathias Hallberg (b), Curt Petterson (a), Mikael Hedeland (a)

[email protected]

(a) Analytical Science, Department of Medicinal Chemistry, Biomedicinskt Centrum, Uppsala universitet. (b) Biological research on drug dependence, Department of Pharmaceutical Biosciences, Uppsala

University

Lipidomics aims to study the lipid profile (lipidome) and the lipid regulation of a biological system. Lipids, which were for long time considered just a barrier of cells, play a key role in physiological and pathophysiological processes. Specifically, the polyunsaturated fatty acids are involved in the triggering and the resolution of inflammation. The state-of-the-art analytical technique in lipidomics separates the lipids by liquid chromatography hyphenated to mass spectrometry. This technique allows the detection of minor species, which is not possible in NMR, and the characterization of the structure of the lipids by fragmentation in tandem mass spectrometry. Here we present a platform to analyze the lipidome. This platform integrates: 1) sample preparation, 2) chromatographic separation, 3) mass spectrometry detection, 4) data pre-treatment, and 5) data statistical analysis. In detail: 1) We adapt the sample preparation to the specific sample type. 2) We perform the chromatographic separation by ultrahigh-performance liquid chromatography in an Acquity chromatographer (Waters) by a gradient on a BEH C18 column. 3) We detect the molecules by mass spectrometry in both negative and positive modes. We ionize the analytes by electrospray (ESI) on a Synapt G2S-Q-ToF analyzer (Waters). 4) We pre-treat the mass spectrometric data in R to identify the lipids by their mass-to-charge and their fragmentation patterns. 5) We perform in R the data analysis plan �which we design at the beginning of the analytic process. Typically, we perform univariate (t-test, ANOVA) and multivariate analysis (principal component analysis, orthogonal partial least squares, elastic nets�). By the application of this workflow, we could identify by multivariate analysis a general shift in the lipidome of blood plasma in rats treated with different anabolic androgenic steroids. Some lipids deviated from this general shift, which were associated to different metabolic pathways. Further studies on these shifts are warranted in order to elucidate their impact on changes in behavior and disease risk. In conclusion, we present a state-of-the-art lipidomics platform to unravel the lipid biology and generate lipid biomarkers of health and disease


30

30

DISCOVERY AND CHARACTERIZATION OF SMALL PEPTIDE TOXINS FROM THE EPIDERMAL MUCUS OF LINEUS LONGISSIMUS, THE LONGEST ANIMAL ON

EARTH

Erik Jacobsson (a), Håkan S. Andersson (b), Malin Strand (c), Steve Peigneur (d), Eline K. M. Lebbe(d), K. Johan Rosengren (e), Jan Tytgat (d), Ulf Göransson (a)

[email protected]

(a) Pharmacognosy, Department of Medicinal Chemistry, Uppsala University, Uppsala, UPPSALA, Sweden (b) Department of Chemistry and Biomedical Sciences, Centre for Biomaterials Chemistry, Linnaeus University, Kalmar, Sweden. (c) Swedish Species Information Centre, Swedish University of Agricultural Sciences, Uppsala,

Sweden (d)Toxicology and Pharmacology, University of Leuven , Leuven, Belgium (e) School of Biomedical Sciences, The University of Queensland, Brisbane, Australia

Peptide toxins display many applications, including use in agriculture, medicine and pharmacology. Some animals have successfully been investigated for peptide toxin presence, while others have been neglected. One neglected source of small peptide toxins is the epidermal mucus secreted by nemertean worms. We have discovered and characterized a novel class of small cystine-rich neurotoxins, a-nemertides, in the mucus of the world’s longest animal: Lineus longissimus. The toxins could only be isolated in minute amounts and were sequenced using mass spectrometry in combination with transcriptomic data. The most abundant toxin, nemertide a-1, was synthetized by solid phase peptide synthesis in sufficient amounts for NMR-based structural elucidation and functional bioassays. Upon injection, nemertide a-1 induces paralysis in Carcinus maenas at a concentration of 300 fmol/kg. In an Artemia salina assay, an IC50 of less than 1 µM was observed. Moreover, nemertide a-1 modulates voltage-gated sodium channels in insects at nM concentrations, while being approximately two orders of magnitude less active in vertebrates. These data indicate a potential use of a-nemertides in insecticidal applications.


31

31

CLINICAL METABOLOMICS: THE STETHOSCOPE FOR THE 21ST CENTURY?

Stephanie Hermana (a,b), Payam Emami Khoonsari (a), Valter Niemelä (c), Torbjörn Åkerfeldta (a), Ola Spjuth (b), Joachim Burman (c), Kim Kultima (a)

[email protected]

(a) Department of Medical Sciences, Clinical Chemistry, Uppsala University, Sweden, (b) Department of Pharmaceutical Biosciences, Uppsala University, Sweden, (c) Department of

Neuroscience, Neurology, Uppsala University, Sweden The metabolome is composed of low-weight molecules or metabolites comprising all biochemical networks in our body. It is the closest omics layer to phenotype and is therefore often called the phenome, and any pathophysiological change occuring in the phenotype will have an immediate impact on the metabolome. As such, the metabolome is an excellent compartment to monitor and measure diversions from normal state. We have phenotyped CSF of different stages of the neurological diseases multiple sclerosis and huntington’s disease as well as control individuals. Both disease metabolomes were significantly different from the control metabolome, with area under the receiver operating characteristic (AUROC) curve values of 0.79(±0.21) and 0.72(±0.14) for multiple sclerosis and huntington’s disease respectively, and demonstrated significant difference between disease stages (AUROC values of 0.83(±0.15) and 0.76(±0.17) respectively). Inter-group comparisons revealed multiple biochemical pathways that were being altered, including the tyrosine metabolism for huntington’s disease, and the tryptophan metabolism and pantothenate and CoA biosynthesis for multiple sclerosis. In addition, the phenylalanine and the nitrogen metabolism were commonly altered in both multiple sclerosis and huntington’s disease. Metabolomics is a relatively young field, which we believe has the potential to develop into something clinically useful. In these studies, the metabolome was able to distinguish the diseases from normal state as well as demonstrated differences between different disease stages.


32

32

SINGLE BEAD INVESTIGATION OF A CLINICAL DRUG DELIVERY SYSTEM

E. Ahnfelt, J. Gernandt, Y. Al-Tikriti, E. Sjögren, H. Lennernäs, and P. Hansson [email protected]

Department of Pharmacy, Uppsala University

Microgels, such as polymeric hydrogels, can be used as drug delivery devices (DDS) for chemotherapeutics and/or unstable drugs. The clinical microgel DC bead® was studied with respect to loading and release, measured as relative bead-volume, of six amphiphilic molecules in a micropipette-assisted microscopy method. Theoretical models for release was used to increase the mechanistic understanding of the DDS. It was shown that loading was independent of amphiphile concentration. Calculations with the developed and applied release model on the release kinetics were consistent with the observations from the micropipette-assisted microscopy method. It was shown that salt was necessary for the release to happen, but at physiological concentrations it was not rate-determining. The rate determining step of the release was the diffusion of the amphiphile monomer through the developed amphiphile aggregate-free depletion layer. The release rate is determined by the diffusivity and the tendency for aggregation of the amphiphile where a weak tendency for aggregation lead to faster release. This highlights the importance of drug self-aggregation on the release rate from this DDS.


33

33

CLASSIFICATION OF DRUG INDUCED TOXICOLOGY IN TISSUE SAMPLES USING MALDI MASS SPECTROMETRY IMAGING

A. Nilsson (a), R.J. Goodwin (b), B. Forngren (c), S. Iverson (c), J. Lindberg (c),

P.E. Andrén (a) [email protected]

(a) Medical Mass Spectrometry, Pharmaceutical Biosciences, UU, (b) AstraZeneca, Cambridge, UK, (c)

former AstraZeneca, Södertälje, Sweden

Traditionally, classical histology has been the authoritative methodology used for the classification and characterization of tissue samples inspected following drug induced toxicological events. However, recent advancements in mass spectrometry imaging (MSI) are providing compelling complementary alternative assays that expand the information obtainable from the analysis of tissue sections. Histology alone can only characterize the changes in tissue morphology and not provide any insight into the underlying molecular interactions occurring within the tissue. Here, we describe how the MS imaging technique can be applied to the identification and monitoring of molecular events behind the toxicology findings in kidneys that represented less pronounced morphological damage, hence difficult to classify by traditional histology. The study included 6 groups of Wistar rats, 3 control groups and 3 groups administered with a cytokine receptor antagonist given as a single dose with necropsy 8h and 24h after dose or repeated dosing for 2 weeks before necropsy 24 hours after the last dose. The caudal poles of the kidneys were snap frozen on dry-ice and used for MALDI MSI analysis and for preparation of kidney extracts. The cranial poles of the kidney were fixed and histologically stained. Histologically stained sections from all animals were graded on a severity scale from 0-5 (5-severe) in terms of tubular casts, tubular simple dilatation, tubular concrement, and tubular degeneration and regeneration. The histopathological examinations revealed no damage 8 hours post administration. At 24 hours, three of six animals showed some kind of damage in terms of tubular casts and one displayed evidence of tubular concrements. All animals in the group that were dosed daily for 2 weeks showed different degrees of kidney damage. Tissue sections, coated with matrix and analyzed at a spatial resolution of 100 µm, were H&E stained post MALDI MS imaging analysis to better reveal the exact concrement location. These stained sections were then aligned to the MALDI MSI results allowing to find molecular ion images that match the pattern of concrement distribution. Several m/z values were found to co-localize to these positions. These m/z values matched very well with metabolites of the administered compound, that previously have been found in plasma, urine, and/or faeces. In total, a subset of four metabolites and some metabolite derived dimers were found to co-localize to concrements. Their identity were further confirmed by MS/MS directly off tissue sections. In addition to identification of drug metabolites at the same localization as the concrements, the MALDI MSI analysis also revealed several potential biomarkers of kidney damage that were localized to the wedge-shaped areas that were described with morphological alterations.


34

34

SCILIFELAB DRUG DISCOVERY AND DEVELOPMENT PLATFORM- MEDICINAL CHEMISTRY

U. Yngve (a) and Y. Gravenfors (b)

[email protected]

(a) SciLifeLab DDD, Medicinal Chemistry-Lead Identification, Department of Medicinal Chemistry, Uppsala University, (b) SciLifeLab DDD, Medicinal Chemistry- Hit to Lead, Department of Organic

Chemistry, Stockholm University

Presentation of the Medicinal Chemistry Facilities at the SciLifeLab Drug Discovery Platform


35

35

ELEVATED BRAIN LEVELS OF L-DOPA AND STRUCTURE SPECIFIC CHANGES IN DOPAMINE METABOLISM IN L-DOPA INDUCED DYSKINESIA REVEALED BY

MASS SPECTROMETRY IMAGING

Elva Friðjónsdóttir (a), Mohammadreza Shariatorji (a), Anna Nilsson (a), Theodosia Vallianatou (a), Per Svenningsson (b), Erwan Bezard (c), Per E. Andrén (a)

[email protected]

(a) Uppsala Univeristy, Uppsala, Sweden, (b) Karolinska Institutet, Stockholm, Sweden, (c) University of Bordeaux, Bordeaux, France

In the early stages of Parkinson’s disease (PD), drug treatment is highly effective, but with disease progression motor complications become increasingly evident. These complications include involuntary movements (dyskinesia), which can become treatment-limiting. At least half of all patients treated with L-DOPA (the dopamine replacing therapy) develop L-DOPA-induced dyskinesia (LID). There is therefore a great need to define the molecular changes associated with patients that develop LID and, likewise, to discover the molecular mechanisms in those patients not developing dyskinesia in PD. Elucidating the metabolic pattern of L-DOPA in dyskinetic and non-dyskinetic individuals may reveal new information about the underlying cause of LID. In this study we analyzed a unique brain biobank from MPTP parkinsonian primates chronically treated with L-DOPA. Dyskinetic and non-dyskinetic animals were compared. Coronal brain 12 µm sections were placed onto conductive ITO glass slides. Our recently developed method for matrix assisted laser desorption ionization (MALDI) mass spectrometry imaging (MSI) enables quantitative mapping of multiple neurotransmitters and metabolites in a single experiment. The method involves using a charge tagging derivatization step that targets monoamine and phenol groups significantly increasing the sensitivity and signal-to-noise of the analytes. We focused on the metabolism of L-DOPA and all downstream catecholamines along with GABA and 5-HT, neurotransmitters that have been implicated to have a role in dyskinesia. MALDI-MSI experiments were performed using a MALDI-FTICR (Solarix

7T-2Ω mass spectrometer in positive ionization mode. Statistics were performed using a independent sample mann whitney u test and the relationship between dopamine and its metabolites was evaluated using linear regression in r. Mass spectrometry imaging results revealed a substantial increase in L-DOPA and 3-O-methyldopa in the dyskinetic animals. Relative quantitation revealed structure specific differences between non-dyskinetic and dyskinetic animals in dopamine, 3-methoxytyramine and 5-HT. In addition, we performed an extensive investigation of dopamine distribution within the putamen, caudate, globus pallidus, hippocampus and cortex areas and identified unusual spatial localization of dopamine. These findings bring new insight into how L-DOPA is distributed and metabolized in LID individuals and generate new hypothesis on how L-DOPA is causing dyskinesia.


36

36

A MASS SPECTROMETRY IMAGING APPROACH FOR INVESTIGATING HOW DRUG-DRUG INTERACTIONS INFLUENCE DRUG BLOOD-BRAIN BARRIER

PERMEABILITY

Theodosia Vallianatou (a), Nicole Strittmatter (b), Anna Nilsson (a), Mohammadreza Shariatgorji (a), Gregory Hamm (b), Marcela Pereira (c), Patrik Källback (a), Per Svenningsson (c), Maria Karlgren (d),

Richard J.A. Goodwin (b), A. Per E. Andrén (a) [email protected]

(a) Medical Mass Spectrometry Imaging, Science for Life Laboratory, National Resource for Mass

Spectrometry Imaging, Department of Pharmaceutical Biosciences, Uppsala University, Box 591, 75124, Uppsala, Sweden, (b) AstraZeneca, Drug Safety and Metabolism, Cambridge, CB4 0WG, UK, (c)

Department of Neurology and Clinical Neuroscience, Karolinska Institutet, 17176, Stockholm, Sweden, (d) Department of Pharmacy, Uppsala University, BMC, Box 580, Uppsala SE-751 23, Sweden.

Several imaging techniques have recently emerged an important role in the investigation of the brain distribution and the blood brain barrier permeability of various compounds. In the present study, we have developed a mass spectrometry imaging (MSI) technique to visualize and quantify the brain distribution of drugs with varying blood-brain barrier permeability. From this perspective, propranolol, a highly permeable compound, was administered intraperitoneally (i.p.) at a dose of 10 mg/kg in mice (n=3) and loperamide, an MDR1 substrate, was administered i.p. at a dose of 5 mg/kg in mice (n=3). Propranolol was distributed throughout the brain and quantified in various brain structures. On the other hand, loperamide was localized in the choroid plexus of the brain ventricles, showing significantly limited uptake in other brain regions. We have also extended the technique to investigations of drug-drug interactions by inhibiting the membrane transporter multidrug resistance 1 protein (MDR1) with elacridar. Co-administration of elacridar (5 mg/kg, i.v.) significantly affected loperamide’s brain uptake (up to 5-fold in certain brain regions), while having no effect on propranolol. In the present study, we show that our MSI technique enables detailed characterization of drug distribution in the brain and localization of drug molecules with varying brain permeability. Compared to traditional imaging approaches, the developed MSI method offers considerably higher spatial resolution and the ability to provide simultaneous images of several molecular species, serving, therefore, as valuable alternative in early ADME studies. We propose that the described approach can serve as a valuable analytical tool during the development of neuroactive drugs. Grant support: This work was supported by the Swedish Research Council (Medicine and Health #2013-3105, Natural and Engineering Science #2014-6215), the Swedish Brain Foundation, the Swedish Foundation for Strategic Research #RIF14-0078, Science for Life Laboratory Grant, ARIADME (European FP7 ITN Community’s Seventh Framework Program, grant no. 607517).


37

37

ANABOLIC ANDROGENIC STEROIDS INDUCE TOXIC EFFECTS IN RAT PRIMARY CORTICAL CELL CULTURES

Sofia Zelleroth, Erik Nylander, Fred Nyberg, Alfhild Grönbladh, Mathias Hallberg.

[email protected]

The Beijer Laboratory, Biological research on drug dependence, Department of Pharmaceutical Biosciences, Uppsala University

Testosterone and other synthetic anabolic androgenic steroids (AASs) are commonly abused in order to enhance athletic performance and physical appearance. The misuse among non-athletes is a growing public health-problem, as abusers underestimate the negative effects associated with these drugs. In addition to physical adverse effects, AASs are known to affect the central nervous system, and induce diverse psychological effects. The present study investigated the toxicity of four structurally diverse AAS, with the aim to understand how AAS abuse affects the brain. The toxicity of testosterone, nandrolone, stanozolol, and trenbolone was investigated by using mixed primary cortical cell cultures from embryonic rats. The cells were exposed to increasing concentrations of AAS every 24 hour for 3 days. The cells were co-treated with the androgen receptor antagonist, flutamide, to investigate whether the effects were mediated by the androgen receptor. Cellular toxicity was determined by measuring mitochondrial activity, lactate dehydrogenase release and caspase activity. All AASs induced toxic effects in the primary cortical cells in a dose-dependent manner. Testosterone, nandrolone, and trenbolone induced apoptotic cell death, in contrast to stanozolol that probably caused the toxic effects by induction of necrosis. Flutamide were able to abolish or diminish the AAS-induced toxicity. In conclusion, supra-physiological concentrations of the AASs induce neurotoxic effects, but to different extent and possibly by different mechanisms. However, the main effects seem to be mediated by activation of the classical AR-pathway.


38

38

CAUSAL FACTORS FOR POST-CRYOPRESERVATION VARIABILITY IN ATTACHMENT AND CULTURE MORPHOLOGY OF PRIMARY HUMAN

HEPATOCYTES

M. Ölander (a), J.R. Wisniewski (b), I. Flörkemeier (a), A. Norén (c), and P. Artursson (a) [email protected]

(a) Department of Pharmacy, Uppsala University, Uppsala, Sweden, (b) Biochemical Proteomics Group, Department of Proteomics and Signal Transduction, Max Planck Institute of Biochemistry, Martinsried,

Germany, (c) Department of Surgical Sciences, Uppsala University, Uppsala, Sweden

Isolated human hepatocytes are used in all aspects of liver research, from the clinical management of liver failure to in vitro studies of drug disposition. Cryopreservation ensures a steady supply of hepatocytes, but causes a variable and unpredictable loss of the ability to form cell-cell and cell-matrix connections. To identify pathways that could be perturbed for mitigating this problem at the post-thawing stage, we used label-free quantitative global proteomics to investigate differences between attached and non-attached fractions of human hepatocyte batches. Hepatocytes that were unable to attach to a collagen matrix displayed many signs of cellular stress, ultimately leading to apoptosis activation. Further analysis showed increases in early apoptosis immediately after thawing, hinting at the possibility of stress reversal. Therefore, we treated the cells with different compounds aimed at decreasing cellular stress. The most effective compound successfully restored attachment ability and promoted a differentiated morphology with improved metabolic function. Further, the treatment did not result in any acute undesirable shifts in hepatocyte protein expression. In conclusion, we have developed a simple restoration approach for cryopreserved human hepatocytes of suboptimal quality, substantially improving key morphological and functional properties of these cells, and markedly increasing the availability of well-functioning human hepatocytes.


39

39

COMPLETE KNOCKOUT OF ENDOGENOUS MDR1 (ABCB1) IN MDCK WILDTYPE AND MDCK-HMDR1 CELLS BY CRISPR-CAS9

M. Karlgren (a, b), I. Simoff (a), M. Backlund (a, b), C. Wegler (a, c), M. Keiser (d), N. Handin (a), J. Müller

(d), P. Lundquist (a), A-C. Jareborg (e), S. Oswald (d), P. Artursson (a, b) [email protected]

(a) Department of Pharmacy, Uppsala University, Uppsala, Sweden, (b) Uppsala University Drug

Optimization and Pharmaceutical Profiling Platform (UDOPP), Science for Life Laboratory, Uppsala, Sweden, (c) Cardiovascular Metabolic Diseases DMPK, AstraZeneca R&D, Mölndal, Sweden, (d)

Department of Clinical Pharmacology, Center of Drug Absorption and Transport (C_DAT), University Medicine of Greifswald, Germany, (e) Department of immunology, genetics and pathology, Science for

Life Laboratory, Uppsala University, Uppsala, Sweden

Madin-Darby canine kidney (MDCK) II cells transfected with transport proteins are commonly used for studying drug transport. However, endogenous transporters e.g. canine MDR1 (ABCB1) complicates the interpretation of transport studies. The aim of this study was to establish MDCK cell lines lacking endogenous canine MDR1 (cMDR1) expression using the CRISPR-Cas9 gene editing technology. MDCK wildtype cells and MDCK cells stably transfected and overexpressing human MDR1 (hMDR1) were transfected simultaneously with three CRISPR-Cas9 vectors targeting three different regions of the canine ABCB1 gene. After clonal selection and expansion, clones with complete cMDR1 knockout were identified by genomic screening. MDR1 protein was quantified and function was tested using Transwell© experiments and known probe substrates of MDR1. The genetic analysis revealed that clear homozygous indels occurred in five wildtype and in four MDCK-hMDR1 based clones. One wildtype based clone and one MDCK-hMDR1 based clone, both having a four-nucleotide deletion in exon 4 leading to a frameshift and a truncated protein, were selected for further studies. Protein quantification verified complete knock-out of cMDR1 and no compensatory up regulation of other efflux transporters could be detected. Functional studies using known MDR1 substrates verified absence of cMDR1 activity in the wildtype based clone with efflux ratios around 1 and preserved function of the hMDR1 transporter with efflux ratios ranging from 5 to 360 in the MDCK-hMDR1 based clone. In conclusion, using CRISPR-Cas9 canine MDR1 has been successfully knocked out in both a MDCK wildtype and in a MDCK-hMDR1 cell line. These two new MDCK cell lines, both with identical deletions in the canine ABCB1 gene and lack of cMDR1 expression represent excellent in vitro tools for use in drug discovery.


40

40

TARGETED BRAIN DELIVERY OF METHOTREXATE BY GLUTATHIONE PEGYLATED LIPOSOMES: HOW CAN THE FORMULATION MAKE A DIFFERENCE?

Yang Hu (a), Pieter J. Gaillard (b), Elizabeth C.M. de Lange (c), and Margareta Hammarlund-Udenaes

(a) [email protected]

(a) Translational PKPD Research Group, Department of Pharmaceutical Biosciences, Associate Member

of SciLife Lab, Uppsala University, Uppsala, Sweden, (b) 2-BBB Medicines B.V., J.H. Oortweg 19, 2333CH, Leiden, The Netherlands, (c) Predictive Pharmacology Group, Division of Systems Biomedicine and

Pharmacology, Leiden Academic Centre for Drug Research, Leiden University, Leiden, The Netherlands

Glutathione (GSH) can serve as a targeting ligand coupled to PEGylated (PEG) liposomes to safely enhance drug delivery into the brain. However, GSH-coating does not necessarily result in increased brain delivery compared to non-targeted PEG liposomes. One possible factor influencing brain-targeted delivery is the liposomal formulation combined with GSH. The purpose of this study was to quantitatively investigate how conjugation of GSH to different liposomal formulations influence the brain delivery of methotrexate (MTX) in rats. GSH-PEG liposomal MTX based on hydrogenated soy phosphatidylcholine (HSPC) or egg yolk phosphatidylcholine (EYPC) and their corresponding PEG control liposomes were prepared. The brain delivery of MTX after intravenously administering free MTX, four liposomal formulations or free MTX + empty GSH-PEG-HSPC liposomes was evaluated by performing microdialysis in brain interstitial fluid and blood. Compared to free MTX with a steady-state unbound brain-to-plasma concentration ratio (Kp,uu) of 0.10, PEG-HSPC liposomes did not affect the brain uptake of MTX, while PEG-EYPC liposomes improved the uptake (Kp,uu 1.5, p < 0.05). Compared to PEG control formulations, GSH-PEG-HSPC liposomes increased brain delivery of MTX by 4-fold (Kp,uu 0.82, p < 0.05), while GSH-coating on PEG-EYPC liposomes did not result in a further enhancement in uptake. The co-administration of empty GSH-PEG-HSPC liposomes with free MTX did not influence the uptake of MTX into the brain. This work showed that the brain-targeting effect of GSH-PEG liposomal MTX is highly dependent on the liposomal formulation that is combined with GSH, providing insights on formulation optimization of this promising brain delivery platform.


41

41

MAPPING THE BRAIN NEUROTRANSMITTER NETWORK WITH MALDI MASS SPECTROMETRY IMAGING

M. Shariatgorji (a), A. Nilsson (a), L. Odell (b), E. Fridjonsdottir (a), T. Vallianatou (a), L. Katan (b), J.

Sävmarker (b), E. Bezard (c), P. Svenningsson (d), P. E. Andren (a) [email protected]

(a) Medical Mass Spectrometry Imaging, National Resource for Mass Spectrometry Imaging, Science for

Life Laboratory, Department of Pharmaceutical Biosciences, Uppsala University, Box 591, SE-75124 Uppsala, Sweden, (b) Organic Pharmaceutical Chemistry, Department of Medicinal Chemistry, Uppsala

Biomedical Center, Uppsala University, Box 574, SE-75123 Uppsala, Sweden, (c) Université de Bordeaux, Institut des Maladies Neurodégénératives, UMR 5293, Bordeaux, France, , (d) Center for Molecular Medicine, Department of Neurology and Clinical Neuroscience, Karolinska Institutet and Karolinska

University Hospital, 17176 Stockholm, Sweden Small-molecule neurotransmitters, their precursors and metabolites are involved in brain chemical networking to transmit signals between neurons. We have recently introduced a MALDI-MS reactive matrix, which selectively targets primary amine groups on neurotransmitters (Neuron, 2014; 84: 697) to facilitate their mass spectrometry imaging. However, using such reactive matrix to study the molecular pathways is not possible due to its limitation to target all downstream metabolites. We here introduce a novel approach to image a large number of different neurotransmitters, their precursors and metabolites simultaneously from brain tissue sections. Since the majority of small molecule neurotransmitters possess phenolic hydroxyls and/or primary or secondary amines (nucleophilic groups), we have focused on the development of reactive matrices based on a nucleophilic aromatic substitution reaction of the 2-fluoro-1-methyl pyridinium (FMP) cation with these groups. An efficient reactive matrix was designed and synthesized to incorporate two functional domains; a highly reactive but selective electrophilic FMP moiety to facilitate covalent analyte charge-tagging and a conjugated chromophore domain with high absorptivity at typical wavelength of Nd:YAG laser to promote efficient laser desorption. Frozen brain tissues were cut and transferred onto conductive glass slides and reactive matrix solution was sprayed over the tissue sections. MALDI-MSI experiments were performed using a MALDI-FTICR mass spectrometers in positive ionization mode. Without modification, most neurotransmitters are not directly detectable by MALDI-MSI due to their poor ionization efficiency. Developed reactive matrix reacts with neurotransmitters carrying phenolic hydroxyls, primary and secondary amino groups. Resulting charged derivatives have laser desorption and ionization efficiencies sufficient to enable the high sensitivity detection by MALDI-MSI. Using the derivatization approach we were able to map numerous neurotransmitters, such as dopamine and serotonin, their precursors and most of their metabolites directly in brain tissue sections.


42

42

THE INNATE/ADAPTIVE IMMUNE RESPONSE TRIGGERED UPON LOCAL IMMUNOTHERAPY OF ORTHOTOPICALLY GROWING BLADDER CANCER

TUMORS

Iliana K. Kerzeli, Andrea Kurtinovic, Aikaterini Nasi, Ivan Stepánek, Martin Lord, Sara M. Mangsbo [email protected]

Department of Pharmaceutical Biosciences, Science for Life Laboratory, Uppsala University, Uppsala,

Sweden The mouse bladder 49 (MB49) syngeneic tumors respond to various immunotherapies, both stimulatory immunotherapies and check-point blockers; when tumors are cured it is often an anti-tumor memory response established (previous publications by our group and others). Herein we present data suggesting that bladders of orthotopic MB49 tumor bearing mice hold a reservoir of lymphocytes surrounding the tumor, wherein they migrate rapidly upon immune adjuvant stimulation. The tumors appear to form an accessible immune microenvironment rich in vessels that mediate the homing of leukocytes to the tumor site. This is in contrast to MB49 tumors that grow subcutaneously and do not appear to provide the same rapid immune cell infiltration capacity. Herein we further examined the immune cell populations of tumors locally treated with CpG ODN 1668, a type B CpG and a strong immunostimulatory agent or dilution buffer, as a control. Interestingly, we identified markers of high-endothelial venules (HEVs) in the vessels present in tumors from both untreated and treated animals. Thus we conclude that HEVs develop naturally in the orthotopic MB49 model and we wished to explore whether manipulation of the tumor microenvironment (TME) can induce HEVs and how this stimuli affect the infiltration of immune cells. Immunohistochemical profiling revealed that CD11c+ dendritic cells, CCL21+ cells along with B and T lymphocytes are present in various ways in different zones of the tumor with HEVs and CD31+ vessels. A correlation with vessel density and the presence of HEVs was identified in untreated animals. In immune-stimulated tumors, this correlation was lost in the early phase post stimuli. We could also identify that CpG treated bladder displayed increased vascularization around the tumor 24 hours post treatment. Interestingly it appeared as CCL21 expression intratumorally aligned with increased recruitment of CD11c+ DCs in the TME. Intratumorally it was a clear correlation between CCL21 expression with both CD11c+ and CD8+ cell infiltration. As CD11c+ dendritic cells migrate towards gradients of CCL21 via their CCR7 receptor, enabling antigen presentation, it is plausible that antigen-presentation takes place in the TME in relation to this finding. By this preliminary data we conclude that in the orthotopic MB49 tumor model, CCL21 appears to play a role in the induction of adaptive immune responses through the CCL21/CCR7 axis. The role of HEVs, or alternatively lymphatic vessels, for this recruitment is yet to be determined. Currently we are investigating multiplex analyses on bladder tumors, urine and blood from tumor bearing animals to identify cytokines in the TME or systemically that are correlated the formation of an adaptive immune response. Our future goal is to examine what is the key to raise an effective immune response to the so called immune "hot" orthotopic MB49 tumor.


43

43

THE UPPSALA UNIVERSITY DRUG OPTIMIZATION AND PHARMACEUTICAL PROFILING FACILITY (UDOPP) — EMERGING NATIONAL INFRASTRUCTURE

FOR ADME INVESTIGATIONS IN SWEDEN

Pawel Baranczewski, Maria Backlund, Maria Karlgren, Shibu Krishnan, Annika Lindqvist, Maria Mastej, Aljona Saleh, Ivailo Simoff, Richard Svensson and Per Artursson

[email protected]

The Uppsala University Drug Optimization and Pharmaceutical Profiling Facility, Drug Deliver Group, Department of Pharmacy, Uppsala University, BMC, Uppsala, Sweden.

The identification of drug candidates suitable for pre-clinical and clinical development involves not only obtaining sufficient potency on the relevant biological pathway, but also ensuring that the compound’s ADME (absorption, distribution, metabolism, excretion) properties are consistent with obtaining efficacy in vivo. The specific UDOPP goals includes to optimize the ADME and pharmaceutical profile of hits and lead candidates through collaborative research programs together with academic/non-for-profit research groups as well as pharmaceutical and biotech companies, and to provide expert competence within drug optimization and pharmaceutical profiling. UDOPP is part of Science for Life Laboratory Drug Discovery and Development Platform (DDD-P), which contributes to academic research at many universities across Sweden. UDOPP became part of the Innovative Medicine Imitative (IMI), European Gram Negative Antibacterial Engine (ENABLE) consortium in February 2014. UDOPP is responsible for in vitro ADME and pharmaceutical profiling investigations within the consortium. In addition, an important part of UDOPP work is the research and development of new predictive techniques and technologies aimed to help support the drug discovery and development process within the ADME field.


44

44

METABOLIC CONTROL OF DENDRITIC CELL IMMUNOGENICITY

Aikaterini Nasi (a.b), Vishnu Priya Bollampalli(b), Meng Sun(c), Yang Chen(c), Sylvie Amu(b), Susanne Nylén(b), Liv Eidsmo(c), Antonio Gigliotti Rothfuchs(b), Bence Réthi (c)

[email protected]

(a) Department of Pharmaceutical Biosciences, Uppsala University, (b) Department of Microbiology, Tumor and Cell Biology, Karolinska Institutet, (c) Department of Medicine Karolinska Institutet

We have shown that human monocyte-derived dendritic cells (DCs) acquired different characteristics in dense or sparse cultures. Sparse cultures promoted IL-12 producing migratory DCs, whereas dense cultures increased IL-10 production. We show here that differentiation of BMDCs in sparse cultures was essential for acquisition of several key functions desired for tumor-targeting DC vaccines, including mobility to draining lymph nodes, the recruitment and massive proliferation of antigen-specific CD4+ T cells, followed by TH1 polarization. Transcription analyses confirmed higher commitment in sparse cultures towards T cell activation and the up-regulation of fatty acid and cholesterol biosynthesis pathways, suggesting a link between DC immunogenicity and lipid homeostasis regulation.


45

45

STABILIZED CYCLIC PEPTIDES AS SCAVENGERS OF AUTOANTIBODIES: NEUTRALIZATION OF ANTI-CITRULLINATED PROTEIN/PEPTIDE ANTIBODIES

IN RHEUMATOID ARTHRITIS

Sunithi Gunasekera (a), Cátia Fernandes-Cerqueira (b), Stefan Wennmalm (c), Heidi Wähämaa (a), Yngve Sommarin (d), Anca I. Catrina (b), Per-Johan Jakobsson (b), Ulf Göransson (a)

[email protected]

(a) Department of Medicinal Chemistry, Uppsala University, Biomedical Centre, Box 574, SE-75123 Uppsala, Sweden, (b) Rheumatology Unit, Department of Medicine, Karolinska Institutet, Karolinska

University Hospital, Rheumatology Clinic D2:01, 171 76 Solna, Stockholm, Sweden, (c) Royal Institute of Technology-KTH, Applied Physics, Experimental Biomolecular Physics, Scilifelab, Tomtebodavägen 23,

171 6, (d) Euro-Diagnostica AB, Malmö, Sweden

The occurrence of autoantibodies is a hallmark of rheumatoid arthritis, specifically those autoantibodies targeting proteins containing the arginine derived amino acid citrulline. There is strong evidence showing that the occurrence of anti-citrullinated protein/peptide antibodies (ACPA) are involved in disease progression, and ACPA was recently shown to induce pain in animals. Here, we explore a novel concept useful for research, diagnostics, and possibly therapy of autoimmune diseases, namely to directly target and neutralize autoantibodies using peptide binders. A high affinity peptide based scavenger of ACPA was developed by grafting a citrullinated epitope derived from human fibrinogen into a naturally occurring stable peptide scaffold. The best scavenger comprises the truncated epitope alfa-fibrinogen, [Cit573]fib(566-580), grafted into the scaffold sunflower trypsin inhibitor-1, SFTI-1. The final peptide demonstrates low nano molar apparent affinity and superior stability.


46

46

PHOSPHOLIPID BUT NOT NEUTRAL LIPID CONTENT IS A MAJOR DETERMINANT OF INTRACELLULAR BINDING OF DRUGS

Andrea Treyer (a), André Mateus (a), Hinnerk Boriss (b), Pär Matsson (a), Per Artursson (a)

[email protected]

(a) Department of Pharmacy, Uppsala University, Uppsala, Sweden, (b) Sovicell GmbH, Leipzig, Germany

Intracellular unbound drug concentration (Cu,cell) drives pharmacological and toxicological responses for targets inside cells. However, the mechanisms that drive cellular drug binding are not so well understood. The aim of this study was to evaluate drug binding to two cellular components - neutral lipids and phospholipids. Binding to cell homogenates (fu,cell) was measured using equilibrium dialysis for 25 compounds. To study the role of neutral- and phospholipids on drug binding, the 3T3-L1 fibroblast cell line was induced to accumulate each of these components. For neutral lipids, a steatotic, adipocyte-like phenotype was obtained via a modified cell culture protocol. For phospholipids, the cells were exposed to drugs that cause drug-induced phospholipidosis. In addition to fluorescent labelling of the lipids, quantification of phospholipids was performed using the Wako Phospholipids C assay. Affinity to pure phosphatidylcholine was measured using the TRANSIL membrane affinity kit. The fluorescent signal of neutral lipid content increased 8-fold in the neutral lipid model. However, this large difference in neutral lipid content was not reflected in the drug binding results. Total phospholipid content was found to increase 1.5-fold - corresponding to the average change of fu,cell which was 1.7-fold. Binding to pure phospholipid showed a strong correlation with fu,cell. Our results suggest that phospholipid content is a major determinant of drug binding to cell homogenates. This finding could result in improved predictions on drug disposition at the target site and accumulation of drugs in tissues, where lipid content is known.


47

47

THE PROTEOME OF PRIMARY HUMAN HEPATOCYTE SPHEROIDS CULTURED AND USED IN HIGH- THROUGHPUT SCREENING FORMAT

Niklas HANDIN (a), Wojciech SENKOWSKI (b), Magnus ÖLANDER (a), Frida NYBERG (b), Mårten

FRYKNÄS (b), Per ARTURSSON (a) [email protected]

(a) Department of Pharmacy, Uppsala University, Sweden

(b) Department of Medical Sciences, Division of Cancer Pharmacology and Computational Medicine, Uppsala University, Sweden

The liver is responsible for synthesis, metabolism and the detoxification of many substances. Approximately 80% of the total liver volume consists of a cell type called hepatocytes1. Primary human hepatocytes (PHH) are routinely used as the golden standards in studies of drug disposition and toxicity in vitro. However, PHH in standard 2D cultures tends to dedifferentiate and therefore lose their hepatocyte specific properties. Numerus different culture conditions have been developed to prolong the time that hepatocytes keep their hepatocyte specific properties2.Here we addressed these problems by cultivating in a 3D model called spheroids. We decided to culture the PHH spheroids in a 384 well format to enable a smooth transition to High-Throughput Screening (HTS). Several different culture conditions was investigated and validated against freshly isolated human hepatocytes using global quantitative proteomics. After optimization, the spheroids displayed excellent spheroid morphology and cell health. Some pathways, for instance lipid processing, histone protein expression and coagulation protein synthesis, was enriched. However, the majority of the drug transporting and metabolizing enzymes remained comparable to that of freshly isolated PHH for up to 3 weeks in culture.

1. Kmieć, Z. Cooperation of liver cells in health and disease. Adv. Anat. Embryol. Cell Biol. 161, III�XIII, 1-151 (2001). 2. Godoy, P. et al. Recent advances in 2D and 3D in vitro systems using primary hepatocytes, alternative hepatocyte sources and non-parenchymal liver cells and their use in investigating mechanisms of hepatotoxicity, cell signaling and ADME. Arch. Toxicol. 87, 1315�1530 (2013).


48

48

TRANSLATING LONG-READ SEQUENCING TO THE CLINIC: A CASE STUDY USING BCR-ABL1 MUTATION SCREENING IN CHRONIC MYELOID LEUKEMIA

Wesley Schaal (a), Adam Ameur (b,c,) Ulla Olsson-Strömberg (d), Lucia Cavelier (d), Ola Spjuth (a,c)

[email protected]

(a) Department of Pharmaceutical Biosciences, Uppsala University, (b) Department of Immunology, Genetics and Pathology, Uppsala University, (c) Science for Life Laboratory, Uppsala University, (d)

Department of Medical Sciences, Uppsala University Hospital We have evaluated PacBio LR-SMS as a routine clinical screening of BCR-ABL1 resistance mutations in patients suffering from Chronic Myeloid Leukemia (CML), an assay previously run using Sanger sequencing. Common programming tools were used to build an analysis and decision support system which was easy to use for both the technicians uploading data and the clinicians reading the results. This was iteratively designed in concert with the clinicians using the system. A clinical lab running routine analyses of BCL-ABL1 for monitoring patients suffering from Chronic Myeloid Leukemia was successfully converted from Sanger to LR-SMS.


49

49

PALLADIUM(0)-CATALYZED CARBONYLATIVE SYNTHESIS OF N-ACYLSULFONAMIDES USING MOLYBDENUM HEXACARBONYL AS A SOLID

CARBON MONOXIDE SOURCE

Luke S. Schembri, Luke R. Odell [email protected]

Department of Medicinal Chemistry, Uppsala Biomedical Center, Uppsala University, P. O. Box 574, SE-

751 23 Uppsala, Sweden

N-acylsulfonamides represent a useful bioisotere of carboxylic acids that allow for greater molecular elaboration and enhanced hydrogen bonding capabilities. Here, we present a novel and powerful palladium(0)-catalysed carbonylative synthesis of N-acylsulfonamides via attack of an in situ generated sulfonylisocyanate by an electron-rich heterocycle. This study focused on the C-H functionalisation of indoles and pyrroles with aliphatic and aromatic sulfonyl azides to produce 34 indole and pyrrole-substituted N-acylsulfonamides in yields of up to 95%. Conveniently, this method does not require palladium ligands, is compatible with an ex situ, solid CO-source, and employs mild temperatures enabling easy access to this important class of compounds. Acknowledgements: The authors would like to thank Dr Mohammadreza Shariatgorji (Uppsala University) and Dr. Lisa Haigh (Imperial College London, U.K.) for assistance with high-resolution mass determination. The research was supported by Uppsala University.


50

50

TRANSFORMATION OF KR-12 DERIVED CROSS-LINKED CYCLIC DIMERS INTO STABLE AND POTENT ANTIMICROBIAL DRUG LEADS

Taj Muhammad, Adam A. Strömstedt, , Sunithi Gunasekera, Ulf Göransson

[email protected]

Pharmacognosy, Department of Medicinal Chemistry, Uppsala University, Biomedical Centre, Uppsala, Sweden

The human cathelicidin LL-37 is a multifunctional host defense molecule that mediates various host responses including direct antimicrobial activity, chemotaxis, epithelial cell activation, angiogenesis, epithelial wound repair and activation of chemokine secretion. However, LL-37 encounters the generic problems common to most linear counter parts such as short biological half-life, primarily arising from proteolytic susceptibility. A minimalized domain of LL-37 with retained bacteriolytic activity has been identified as KR-12. Based on mapping of essential residues, we have designed cross-linked cyclic analogs to improve peptide stability, while at the same time improve on-target activity. The cross-linked cyclic analogs showed more potent antimicrobial activity against all the tested strains (E. coli, P. aeuroginosa, S. aureus and C. albicans) in biological conditions than the non-cross-linked analogs, activity equivalent to the parent peptide LL-37. In addition to increasing our understanding of the structure activity relationship of KR-12, this study shows that the combining backbone cyclization and inter-helical disulfide bonds is a potency enhancing strategy to improve of the otherwise protease-susceptible KR-12. This peptide-engineering strategy can be a useful tool in peptide antibiotic drug development.


51

51

TIME-DEPENDENT EFFECTS ON SMALL INTESTINAL TRANSPORT BY ABSORPTION-MODIFYING EXCIPIENTS

Dahlgren, David1; Roos, Carl1; Tannergren, Christer4; Lundqvist, Anders4; Sjöblom, Markus2;

Sjögren, Erik1; Lennernäs, Hans1 [email protected]

1Institutionen för Farmaci, UU. 2Institutionen för Neurovetenskap, UU. 3School of Pharmacy, Johannes

Gutenberg-University. 4AstraZeneca, Göteborg. Abstract: The relevance of the rat single-pass intestinal perfusion model for investigating in vivo time-dependent effects of absorption-modifying excipients (AMEs) is not fully established. Therefore, the dynamic effect and recovery of the intestinal mucosa was evaluated based on the lumen-to-blood flux (Jabs) of six model compounds, and the blood-to-lumen clearance of 51Cr-EDTA (CLCr), during and after 15- and 60-min mucosal exposure of the AMEs, sodium dodecyl sulfate (SDS) and chitosan, in separate experiments. The contribution of enteric neurons on the effect of SDS and chitosan was also evaluated by luminal coadministration of the nicotinic receptor antagonist, mecamylamine. The increases in Jabs and CLCr (maximum and total) during the perfusion experiments were dependent on exposure time (15 and 60 min), and the concentration of SDS, but not chitosan. The increases in Jabs and CLCr following the 15-min intestinal exposure of both SDS and chitosan were greater than those reported from an in vivo rat intraintestinal bolus model. However, the effect in the bolus model could be predicted from the increase of Jabs at the end of the 15-min exposure period, where a six-fold increase in Jabs was required for a corresponding effect in the in vivo bolus model. This illustrates that a rapid and robust effect of the AME is crucial to increase the in vivo intestinal absorption rate before the yet unabsorbed drug in lumen has been transported distally in the intestine. Further, the recovery of the intestinal mucosa was complete following 15-min exposures of SDS and chitosan, but it only recovered 50% after the 60-min intestinal exposures. Our study also showed that the luminal exposure of AMEs affected the absorptive model drug transport more than the excretion of 51Cr-EDTA, as Jabs for the drugs was more sensitive than CLCr at detecting dynamic mucosal AME effects, such as response rate and recovery. Finally, there appears to be no nicotinergic neural contribution to the absorption-enhancing effect of SDS and chitosan, as luminal administration of 0.1 mM mecamylamine had no effect.


52

52

EFFECT OF ABSORPTION-MODIFYING EXCIPIENTS, HYPOTONICITY, AND ENTERIC NEURAL ACTIVITY IN AN IN VIVO MODEL FOR SMALL INTESTINAL

TRANSPORT Dahlgren, David1; Roos, Carl1; Tannergren, Christer4; Lundqvist, Anders4; Sjöblom, Markus2;

Sjögren, Erik1; Lennernäs, Hans1 [email protected]

1Institutionen för Farmaci, UU. 2Institutionen för Neurovetenskap, UU. 3School of Pharmacy, Johannes

Gutenberg-University. 4AstraZeneca, Göteborg. The small intestine mucosal barrier is physiologically regulated by the luminal conditions, where intestinal factors, such as diet and luminal tonicity, can affect mucosal permeability. The intestinal barrier may also be affected by absorption-modifying excipients (AME) in oral drug delivery systems. Currently, there is a gap in the understanding of how AMEs interact with the physiological regulation of intestinal electrolyte transport and fluid flux, and epithelial permeability. Therefore, the objective of this single-pass perfusion study in rat was to investigate the effect of three AMEs on the intestinal mucosal permeability at different luminal tonicities (100, 170, and 290 mOsm). The effect was also evaluated following luminal administration of a nicotinic receptor antagonist, mecamylamine, and after intravenous administration of a COX-2 inhibitor, parecoxib, both of which affect the enteric neural activity involved in physiological regulation of intestinal functions. The effect was evaluated by changes in intestinal lumen-to-blood transport of six model compounds, and blood-to-lumen clearance of 51Cr-EDTA (a mucosal barrier marker). Luminal hypotonicity alone increased the intestinal epithelial transport of 51Cr-EDTA. This effect was potentiated by two AMEs (SDS and caprate) and by parecoxib, while it was reduced by mecamylamine. Consequently, the impact of enteric neural activity and luminal conditions may affect nonclinical determinations of intestinal permeability. In vivo predictions based on animal intestinal perfusion models can be improved by considering these effects. The in vivo relevance can be increased by treating rats with a COX-2 inhibitor prior to surgery. This decreases the risk of surgery-induced ileus, which may affect the physiological regulation of mucosal permeability.


53

53

THE EFFECTS OF THREE ABSORPTION-MODIFYING CRITICAL EXCIPIENTS ON THE IN VIVO INTESTINAL ABSORPTION OF SIX MODEL COMPOUNDS IN RATS

AND DOGS

Dahlgren, David1; Roos, Carl; Johansson, Pernilla4; Tannergren, Christer4; Lundqvist, Anders4; Langguth, Peter3; Sjöblom, Markus2; Sjögren, Erik1; Lennernäs, Hans1

[email protected]

1Institutionen för Farmaci, UU. 2Institutionen för Neurovetenskap, UU. 3School of Pharmacy, Johannes Gutenberg-University. 4AstraZeneca, Göteborg.

Pharmaceutical excipients that may affect gastrointestinal (GI) drug absorption are called critical pharmaceutical excipients (CPEs), or absorption-modifying excipients (AMEs) if they act by altering the integrity of the intestinal epithelial cell membrane. Some of these excipients increase intestinal permeability, and subsequently the absorption and bioavailability of the drug. This could have implications for both the assessment of bioequivalence and the efficacy of the absorption-enhancing drug delivery system. The absorption-enhancing effects of AMEs/CPEs with different mechanisms (chitosan, sodium caprate, sodium dodecyl sulfate (SDS)) have previously been evaluated in the rat single-pass intestinal perfusion (SPIP) model. However, it remains unclear whether these SPIP data are predictive in a more in vivo like model. The same excipients were in this study evaluated in rat and dog intraintestinal bolus models. SDS and chitosan did exert an absorption-enhancing effect in both bolus models, but the effect was substantially lower than those observed in the rat SPIP model. This illustrates the complexity of the AME/CPE effects, and indicates that additional GI physiological factors need to be considered in their evaluation. We therefore recommend that AME/CPE evaluations obtained in transit-independent, preclinical permeability models (e.g. Ussing, SPIP) should be verified in animal models better able to predict in vivo relevant GI effects, at multiple excipient concentrations.


54

54

PRECLINICAL EFFECT OF ABSORPTION-MODIFYING EXCIPIENTS ON RAT INTESTINAL TRANSPORT OF MODEL COMPOUNDS AND THE MUCOSAL

BARRIER MARKER 51CR-EDTA

Dahlgren, David1; Roos, Carl1; Tannergren, Christer4; Lundqvist, Anders4; Langguth, Peter3; Sjöblom, Markus2; Sjögren, Erik1; Lennernäs, Hans1

[email protected]

1Institutionen för Farmaci, UU. 2Institutionen för Neurovetenskap, UU. 3School of Pharmacy, Johannes Gutenberg-University. 4AstraZeneca, Göteborg.

There is a renewed interest from the pharmaceutical field to develop oral formulations of compounds such as peptides, oligonucleotides, and polar drugs. However, these often suffer from insufficient absorption across the intestinal mucosal barrier. One approach to circumvent this problem is the use of absorption modifying excipient(s) (AME). This study determined the absorption enhancing effect of four AMEs (sodium dodecyl sulfate, caprate, chitosan, N-acetylcysteine) on five model compounds in a rat jejunal perfusion model. The aim was to correlate the model compound absorption to the blood-to-lumen clearance of the mucosal marker for barrier integrity, 51Cr-EDTA. Sodium dodecyl sulfate and chitosan increased the absorption of the low permeation compounds but had no effect on the high permeation compound, ketoprofen. Caprate and N-acetylcysteine did not affect the absorption of any of the model compounds. The increase in absorption of the model compounds was highly correlated to an increased blood-to-lumen clearance of 51Cr-EDTA, independent of the AME. Thus, 51Cr-EDTA could be used as a general, sensitive, and validated marker molecule for absorption enhancement when developing novel formulations.


55

55

TRANSFER LEARNING WITH DEEP CONVOLUTIONAL NEURAL NETWORKS FOR CLASSIFYING CELLULAR MORPHOLOGICAL CHANGES

Alexander Kensert, Philip J Harrison, Ola Spjuth

[email protected]

Department of Pharmaceutical Biosciences, Uppsala University Quantification and identification of cellular phenotypes from high content microscopy images have proven to be very useful for understanding biological activity in response to different drug treatments [1]. The traditional approach has been to use classical image analysis to quantify changes in cell morphology, which requires several non-trivial and independent analysis steps. Recently convolutional neural networks have emerged as a compelling alternative, offering good predictive performance and the possibility to replace traditional workflows with a single network architecture [2,3]. In this study we applied the pre-trained deep convolutional neural networks ResNet50, InceptionV3 and InceptionResnetV2 to predict cell mechanisms of action in response to chemical perturbations for two cell profiling datasets from the Broad Bioimage Benchmark Collection [4]. These networks were pre-trained on ImageNet enabling much quicker model training. We obtain higher predictive accuracy than previously reported, between 95 and 97%. The ability to quickly and accurately distinguish between different cell morphologies from a scarce amount of labelled data illustrates the combined benefit of transfer learning and deep convolutional neural networks for interrogating cell-based images. ACKNOWLEDGEMENTS The computations were partly performed on resources provided by SNIC through Uppsala Multidisciplinary Center for Advanced Computational Science (UPPMAX). This project received financial support from the Swedish Foundation for Strategic Research (SSF) as part of the HASTE project under the call ’Big Data and Computational Science’. REFERENCES [1] Ljosa, V., Caie, P.D., ter Horst, R., Sokolnicki, K.L., Jenkins, E.L., Daya, S., Roberts, M.E., Jones, T.R., Singh, S., Genovesio, A., Clemons, P.A., Carragher, N.O., Carpenter, A.E., 2013. Comparison of methods for image- based profiling of cellular morphological responses to small-molecule treatment. Journal of Biomolecular Screening 18, 1321–1329. [2] Godinez, W.J., Hossain, I., Lazic, S.E., Davies, J.W., Zhang, X., 2017. A multi-scale convolutional neural network for phenotyping high-content cellular images. Bioinformatics 33, 2010–2019. [3] Pärnamaa, T., Parts, L., 2017. Accurate classification of protein subcellular localization from high-throughput microscopy images using deep learning. G3: Genes, Genomes, Genetics 7, 1385–1392. [4] Vebjorn, L., L, S.K., E, C.A., 2012. Annotated high-throughput microscopy image sets for validation. Nature Methods 9, 637 EP.


56

56

SUCCESSFUL METHOD VALIDATION AND ABSOLUTE QUANTITATION BY MALDI MASS SPECTROMETRY IMAGING

Patrik Källback, Theodosia Vallianatou, Anna Nilsson, Mohammadreza Shariatgorji, Nicoletta Schintu, Marcela Pereira, Florian Barré, Henrik Wadensten, Per Svenningsson and Per E. Andrén

[email protected]

Department of Pharmaceutical Biosciences, Medical Mass Spectrometry Imaging Introduction. The validation of analytical methods has become an essential part of successful drug development and characterization to become widely accepted and must be performed according to the authorities’ guidelines. To establish a quantitative bioanalytical method, following parameters must be validated: selectivity, accuracy, precision, recovery, linearity, measurement range, and stability. It should also be compared by an established reference method and show non-significant deviation. Until now, no complete bioanalytical method validation according to these guidelines has been performed using MALDI-MSI. A few of the pharmaceutical drug quantitative studies that are reported for MALDI-MSI have been confirmed by comparative techniques. Here we report a comprehensive method validation and absolute quantitation protocol for MALDI-MSI, which was cross-validated with LC-MS/MS. Brain tissue from mice dosed by citalopram (CIT) was used as the biological experimental model. We used two different MALDI-MSI instruments, TOF and FTICR that were compared with a validated LC-MS/MS method. Methods. Mice were administrated CIT (dose) intraperitoneally (i.p.). Coronal brain tissue sections were cut using a cryomicrotome at 14 µm thickness. Consecutive sections were collected for either MALDI-TOF-MSI or MALDI-FTICR-MSI and were placed onto ITO-coated glass slides. Tissue sections for LC-MS were collected into Eppendorf tubes. The selectivity was determined by measuring the ratio between lower limit of quantitation (LLOQ) and blank sample. Accuracy, precision and recovery were measured by applying zero samples, standard curves, and quality controls (QCs) spots deposited at 50 nL volume on control tissue. QC were also spotted directly on the glass slide to measure the recovery of analyte and internal standard. The MALDI-target was coated by a uniform layer of internal standard followed by the addition of the MALDI-matrix. The resulting ion intensity distribution of citalopram was normalized to the internal standard. The tissue sections for LC-MS/MS analysis were homogenized and the internal standard was added prior to liquid-liquid extraction and sample filtration. The sample was then evaporated and reconstituted before further analysis. Results. The selectivity is estimated at the LLOQ and is required to be at least 5 times the response compared to the blank sample response. Our selectivity results showed that the ratio between detector response for LLOQ and blank sample for both MALDI-MSI methods were acceptable according to guidelines, MALDI-TOF-MSI had a ratio of 11 and MALDI-FTICR-MSI a ratio of 22. The accuracy and precision were estimated by applying quality controls to six different control tissues. The acceptance

limits of accuracy and precision are ≤ 15 % when measuring QCs except for LLOQ, which has an

acceptance limit of ≤ 20 %. Both MALDI instrument values were within acceptance criteria. MALDI-TOF-MSI had a QC accuracy range: 0.7 – 14.9 % (LLOQ: 5.1 – 17.0) and a QC precision range: 6.9 – 14.5 % (LLOQ: 9.9 – 14.8). MALDI-FTICR-MSI had a QC accuracy range: 3 – 13.6 % (LLOQ: 2.4 – 14.2) and a QC precision range: 5.9 – 12.6 % (LLOQ: 3.8 – 16.4). The cross-validation was performed to investigate whether there were any significant differences of the quantified tissue concentration of CIT between MALDI-MSI and LC-MS/MS. Two tests were conducted in which the method and the biological variance were investigated of the quantified tissue sections. When the method variance and biological variance were compared with LC-MS/MS, neither MALDI-TOF-MSI nor MALDI-FTICR-MSI showed any significant difference. Conclusions. We have performed complete method validation and quantitation for MALDI-MSI that was cross-validated with LC-MS/MS. The method validation were within specified limits for selectivity, accuracy, precision and the quantitation results showed no significant difference when comparing the MALDI-MSI methods against the LC-MS/MS reference method.


57

57

JEJUNAL ABSORPTION OF APREPITANT FROM SUSPENSIONSIN RAT: ROLE OF PARTICLE SIZE, PRANDIAL STATE AND MUCUS LAYER

Carl Roos1, David Dahlgren1, Erik Sjögren1, Markus Sjöblom1, Mikael Hedeland1, 1, Hans

Lennernäs1 [email protected]

1Uppsala University, 2National Veterinary Institute, Uppsala

The number of highly lipophilic active pharmaceutical ingredients (APIs) in pharmaceutical development has been constantly increasing over recent decades. TheseAPIs often have inherent issues with solubility and dissolution, limiting their oral bioavailability. Traditionally, a reduction in particle size to the micrometer range has been used to improve dissolution. More recently, size reduction to the nanometer range has been introduced, which further increases the dissolution rate, but may also involve other mechanisms for increasing bioavailability. The effect of particle size on the absorption of aprepitant was investigated using the single-pass intestinal perfusion (SPIP) model in the rat jejunum. Phosphate buffer, fasted-state simulated intestinal fluid (FaSSIF), and fed-state simulated intestinal fluid (FeSSIF) were used as perfusion media to increase understanding of the processes involved and the effects of colloidal structures. The role of mucus on intestinal absorption was investigated by adding the mucolytic agent N-acetyl-cysteine (NAC). The absorption of aprepitant from the nanosuspensions was similar with all perfusion media (buffer=FaSSIF=FeSSIF), whereas food had a pronounced effect on absorption from the microsuspensions (FeSSIF > FaSSIF > buffer). The colloidal structures hence contributed to absorption from the microsuspensions. Partitioning of aprepitant from the nanosuspension into the colloidal structures decreased the amount of nanoparticles available, which offset the effect of food. The appearance flux of aprepitant in blood was non-significantly decreased for nanosuspensions of aprepitant with NAC versus without NAC in buffer (ratio of 2:1), indicating that particle deposition in the mucus may have been decreased as the layer thinned, with subsequently reduced intestinal absorption. The study also showed that the SPIP model is suitable for investigating detailed absorption mechanisms using complex perfusion media, which increase the biorelevance of the model.


58

58

MECHANISTIC MODELLING OF INTESTINAL DRUG ABSORPTION – THE IN VIVO EFFECTS OF NANOPARTICLES, HYDRODYNAMICS, AND COLLOIDAL

STRUCTURES

Carl Roos1, Jan Westergren2, David Dahlgren1, Hans Lennernäs1, Erik Sjögren1 [email protected]

1Uppsala University, 2 Wendelsbergs beräkningskemi AB, Mölnlycke

Particle size reduction is a traditional approach to increase the intestinal absorption of active pharmaceutical ingredients with poor intestinal solubility, by increasing the particle dissolution rate. However, an increase in the dissolution rate cannot always fully explain the effects of nanoformulations, and a method of assessing the potential benefits of a nanoformulation in vivo would hence be of great value in drug development. A novel mathematical model of a nanoformulation, including interlinked descriptions of the hydrodynamics, particle dissolution and diffusion of particles and colloidal structures (CS), was developed to predict the combined in vivo effects of these mechanisms on drug absorption. The model successfully described previously reported in vivo observations of nanoformulated aprepitant in rats, at various drug concentrations and in the presence or absence of CS. The increase in absorption rate was explained as a direct consequence of the increased drug concentration at the membrane, caused by the contributing effects of the diffusion of both nanoparticles and CS into which the drug had partitioned. Further simulations supported the conclusion that the model can be applied during drug development to provide a priori assessments of the potential benefits of nanoformulations.


59

59

DEVELOPMENT OF SCAFFOLD-BASED PEPTIDE SCAVENGERS FOR INHIBITION OF AUTOANTIBODIES IN RHEUMATOID ARTHRITIS

Camilla Eriksson (a), Sunithi Gunasekera (a), Cátia Fernandes-Cerqueira (b), Per-Johan Jakobsson (b),

Ulf Göransson (a) [email protected]

(a) Uppsala University, Department of Medicinal Chemistry, Division of Pharmacognosy, Biomedical

Centre, Uppsala University, Husargatan 3, 751 23, Uppsala, Sweden. (b) Karolinska University Hospital, Department of Medicine, Rheumatology Unit, Rheumatology Clinic D2:01, 171 76, Stockholm, Sweden.

INTRODUCTION Rheumatoid arthritis (RA) is an inflammatory autoimmune disease characterised by painful deformation and degradation of bone and cartilage in synovial joints. A new class of rheumatoid arthritis-specific autoantibodies has recently been discovered that target citrullinated proteins: the so-called Anti-Citrullinated Protein Antibodies (ACPA)1. Their high specificity for RA2 and early appearance in the disease course makes them important serological markers. Several lines of evidence support the involvement of ACPA in RA pathogenesis3 and it has been hypothesised that inhibition of these autoantibodies could be useful in development of novel therapies4. Using known target sequences and stable cyclic peptide scaffolds with epitope-stabilising properties5,6, we developed a set of citrullinated peptide analogues with in vitro ACPA neutralising activity. MATERIALS AND METHODS Citrullinated peptides and their corresponding arginine-controls were synthesised using FMOC-SPPS, purified by RP-HPLC and analysed by analytical LC-UV and MS. Specifically, scaffold peptides were synthesised by grafting the citrullinated epitope onto the scaffold peptide sunflower trypsin inhibitor 1 (SFTI-1)6. Peptides at different concentrations were incubated with ACPA purified from patient sera. Neutralising activity was assessed by measuring the remaining ACPA reactivity to surrogate peptide targets in CCP-2 ELISA. RESULTS The citrullinated peptide analogues demonstrated autoantibody neutralising activity when compared to arginine controls. In addition, improved neutralising activity was observed for the scaffold peptide analogues in comparison to their corresponding linear peptide epitopes. CONCLUSIONS These results further support the potential of scaffold-based citrullinated peptides as inhibitors of anti-citrullinated protein autoantibodies in rheumatoid arthritis. 1. Schellekens, G.A. et al. (1998) J Clin Invest. 101(1), pp. 273–281. 2. Coenen, D. et al. (2007) Clin Chem. 53 (3): 498-504. 3. Forslind, K. et al. (2004) Ann Rheum Dis 63:1090–1095. 4. Cerqueira, F.C. et al. (2014) Basic Clin. Pharmacol. Toxicol. 114(1), 13-17. 5. Burman, R. et al. (2014) J Nat Prod. 77(3):724-36. 6. Gunasekera, S. et al. (2008) J Med Chem. 51(24):7697-704.

POSTER SESSION - uu-pharmfaculty50years.seuu-pharmfaculty50years.se/wp-content/uploads/2018/10/Poster-session... · i FACULTY OF PHARMACY 50 YEARS POSTER SESSION 26 October 2018 Full

Documents

POSTER SESSION - uu-pharmfaculty50years.seuu-pharmfaculty50years.se/wp-content/uploads/2018/10/Poster-session... · i FACULTY OF PHARMACY 50 YEARS POSTER SESSION 26 October 2018 Full