International review of cell and molecular biology, volume 319

Institute of Microbiology, Technische Universit €at Braunschweig, Braunschweig, Lower Saxony, Germany*Corresponding authors: E-mail: simone.bergmann@tu-bs.de; m.steinert@tu-braunschweig.

VOLUME THREE HUNDRED AND NINETEEN

CELL AND MOLECULAR BIOLOGY

International Review of Cell and Molecular Biology

Series Editors

Editorial Advisory Board

WALLACE F MARSHALL

VOLUME THREE HUNDRED AND NINETEEN

AMSTERDAM • BOSTON • HEIDELBERG • LONDON

NEW YORK • OXFORD • PARIS • SAN DIEGO

SAN FRANCISCO • SINGAPORE • SYDNEY • TOKYO

Academic Press is an imprint of Elsevier

Academic Press is an imprint of Elsevier

225 Wyman Street, Waltham, MA 02451, USA

525 B Street, Suite 1800, San Diego, CA 92101-4495, USA

125 London Wall, London EC2Y 5AS, UK

The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, UK

First edition 2015

Copyright © 2015 Elsevier Inc All rights reserved.

No part of this publication may be reproduced or transmitted in any form or by any means, electronic or mechanical, including photocopying, recording, or any information storage and retrieval system, without permission in writing from the publisher Details on how to seek permission, further information about the Publisher’s permissions policies and our arrangements with organizations such as the Copyright Clearance Center and the Copyright Licensing Agency, can be found at our website: www.elsevier.com/permissions

This book and the individual contributions contained in it are protected under copyright by the Publisher (other than as may be noted herein).

Notices

Knowledge and best practice in this field are constantly changing As new research and experience broaden our understanding, changes in research methods, professional practices,

or medical treatment may become necessary.

Practitioners and researchers must always rely on their own experience and knowledge in evaluating and using any information, methods, compounds, or experiments described herein In using such information or methods they should be mindful of their own safety and the safety of others, including parties for whom they have a professional responsibility.

To the fullest extent of the law, neither the Publisher nor the authors, contributors, or editors, assume any liability for any injury and/or damage to persons or property as a matter of products liability, negligence or otherwise, or from any use or operation of any methods, products, instructions, or ideas contained in the material herein.

ISBN: 978-0-12-802278-8

ISSN: 1937-6448

For information on all Academic Press publications

visit our website at http://store.elsevier.com/

Institute of Microbiology, Technische Universität Braunschweig, Braunschweig,

Lower Saxony, Germany

Eugene Bell Center for Regenerative Biology and Tissue Engineering, Marine Biological Laboratory, Woods Hole, MA; Department of Biology, University of Maryland,

College Park, MD, USA

Kishore R Katikireddy

Schepens Eye Research Institute & Mass Eye and Ear, Department of Ophthalmology, Harvard Medical School, Boston, MA, USA

Ikeda Lal

LV Prasad Eye Institute, Hyderabad, Telangana, India

Jung Weon Lee

Department of Pharmacy, Research Institute of Pharmaceutical Sciences, Tumor

Microenvironment Global Core Research Center, Medicinal Bioconvergence Research Center, College of Pharmacy, Seoul National University, Seoul, Korea

ixj

Dilip Kumar Mishra

Department of Ocular Pathology, LV Prasad Eye Institute, Hyderabad, Telangana, India Purnima Nerurkar

Institute of Biochemistry (IBC), Department of Biology (D-BIOL), ETH Zurich; Molecular Life Science (MLS) Graduate School, Zurich, Switzerland

Vikram Govind Panse

Institute of Biochemistry (IBC), Department of Biology (D-BIOL), ETH Zurich, Zurich, Switzerland

Charanya Ramachandran

Sudhakar and Sreekanth Ravi Stem Cell Biology Laboratory, Prof Brien Holden Eye Research Centre, C-TRACER, LV Prasad Eye Institute, Hyderabad, Telangana, India Eileen G Russell

Tumour Biology Laboratory, School of Biochemistry and Cell Biology, Bioscience Research Institute, University College Cork, Cork, Ireland

Sudhakar and Sreekanth Ravi Stem Cell Biology Laboratory, Prof Brien Holden Eye Research Centre, C-TRACER, LV Prasad Eye Institute, Hyderabad, Telangana, India Michael Steinert

Institute of Microbiology, Technische Universität Braunschweig, Braunschweig,

Lower Saxony, Germany

Christine Weirich

Institute of Biochemistry (IBC), Department of Biology (D-BIOL), ETH Zurich, Zurich, Switzerland

Institute of Microbiology, Technische Universit €at Braunschweig, Braunschweig, Lower Saxony, Germany

*Corresponding authors: E-mail: simone.bergmann@tu-bs.de; m.steinert@tu-braunschweig.de

2.3.1 Coculture-based generation of tissue barriers 11 2.3.2 Coculture of adherent cells and neutrophiles in suspension 13

3.1 Bene fits and Limitations of 3D Scaffold 15

3.1.1 Coculture-based reconstruction of BBB with matrix scaffold 17 3.1.2 Requirements of 3D tissue models generating aireliquid surface 18

3.2 MicrogravitydVariations of 3D Cell Culture Models 19

4.1 Organoids and Tissue Equivalents Providing Complex Cell Systems

“En Miniature”

22

4.3 Integration of Micro fluidic Systems in 2D and 3D Cell Culture 26

5 Concluding Remarks and Future Perspectives 30

ISSN 1937-6448

http://dx.doi.org/10.1016/bs.ircmb.2015.06.003 © 2015 Elsevier Inc.

All rights reserved 1j

The implementation of natural or synthetical scaffolds elevated the model complexity

to the level of three-dimensional cell culture Additionally, several transwell-based cell culture techniques are applied to study bacterial interaction with physiological tissue barriers For keeping highly differentiated phenotype of eukaryotic cells in ex vivo culture conditions, different kinds of microgravity-simulating rotary-wall vessel systems are employed Furthermore, the implementation of micro fluidic pumps enables con- stant nutrient and gas exchange during cell cultivation and allows the investigation

of long-term infection processes The highest level of cell culture complexity is reached

by engineered and explanted tissues which currently pave the way for a more hensive view on microbial pathogenicity mechanisms.

compre-1 INTRODUCTION

Most basic studies on hostepathogen interaction have been focused

on cultured and, frequently, immortalized cell lines or animal experiments(Mizgerd and Skerrett, 2008) The first reports highlighting the suitability

of in vitro cell culture models to study pathogenesis of microorganismswere published in the early 1970s and focused on virusehost cell interac-tions (Todaro et al., 1971) As early as in 1976, Taylor-Robinson (1976)described the use of ciliated tracheal epithelium of animals to studymycoplasma pneumonia infections Since then, in vitro cell culture modelsbecame increasingly popular in infection biology as they combine severaladvantages Compared to animal models they are cost-effective and acces-sible, they allow experimental flexibility including high-throughput plat-forms and they exhibit a high reproducibility Moreover, the vastinnovations in cell biology such as microscopic imaging, genetic, biochem-ical, and immunologic technologies allowed deep insights in host cell re-sponses elicited by microbes This includes the exploitation of host cellcomponents during adherence, invasion, replication, and evasion of patho-gens Meanwhile, several tissue culture collections and companies offer abroad list of different immortalized cell lines and primary cells derivedfrom human and different animal species thereby allowing cell type-specificinvestigations

The central key point in cell culture-based infection biology is the level

of complexity which can be reached by an in vitro cell culture model in gard to differentiation and reactivity, to adequately mimic the situation in acomplex host organism In order to improve the value of data obtained fromcell culture models, the methods have been optimized and adapted to specialscientific questions Many examples derived from different scientific

disciplines approved that simplified models enhance the probability to date crucial or new specific interactions of individual components byexcluding the vast amount of overlaying interactions A prerequisite forthis, however, is that conclusions drawn from model systems take intoaccount the fact that certain properties are not represented in the model.Thus, it is widely accepted that the suitability of a specific cell culture-basedmodel to generate reliable data, which can be superimposed for the situation

eluci-in vivo, has to be validated for every seluci-ingle scientific question

To overcome the limitation of isolated and often immortalized cells,models of higher complexity have been generated during the last decades

of years Figure 1 depicts the increase in model complexity and outlines

Figure 1 Schematic illustration of different cell culture models presented in this review The cell culture models are roughly categorized in two-dimensional (2D) cell culture, three-dimensional (3D) cell culture with scaffolding materials, and organoids and tissue explants The cell culture models are distributed vertically according to the level of complexity, reaching from cellular level to tissue level The transwell models are composed of a two chamber system separated by a porous membrane In most of the applied transwell cell culture systems, different cell types are cultivated on the up- per and the lower site of the membrane Several bloodebrain barrier (BBB) models also implement scaffolding materials such as extracellular matrix (ECM) components for cell culture The rotary vessel culture enabled cell cultivation in microgravity environment.

the structure of this review Beginning with monolayer cultures and zoa-based models as paradigms for simple two-dimensional (2D) cell culture,

proto-we will discuss the benefits and limitations of established cell culture modelsapplied in infection biology Following the line of increasing complexity,several cocultivation techniques, transwell-based tissue models, and culturesystems with implementation of scaffolds will be presented in the secondpart The current highest level in complexity is reached in reconstructionalsystems on the tissue level and includes the generation of tissue aggregates,the cultivation of organoids and the use of organ-specific ex vivo tissue ex-plants Based on selected examples, the advantages and restrictions of thesecomplex three-dimensional (3D) systems will be commented within thelast chapter of this review

2 2D CELL CULTURE

In many experimental setups, the creation of a functional host cell face is sufficient to study initial interactions with bacterial pathogens such asadherence, invasion, and induction of signal transduction processes For de-cades of years 2D cell monolayers grown on solid, impermeable plastic orglass surfaces have been applied as simple and cost-effective strategy toanalyze principle mechanisms of bacterialehost cell interactions Numerous

sur-in vitro cell culture systems have been confirmed as suitable to provide formation about specific bacterial virulence factors involved and also eluci-dated the induction of many fold intracellular processes, such as signalingcascades and cytoskeletal rearrangements on the host side (Table 1).Depending on the physiological niche of the bacteria, pulmonary cellsare cultivated and infected with typical lung pathogens like Legionella pneu-mophila and Streptococcus pneumoniae; gastrointestinal cells are used for infec-tion studies with Helicobacter and Salmonella, and skin fibroblasts are chosenfor infection with causative agents of wound infections like staphylo-coccidjust giving a few examples These studies generated impressive trans-mission electron microscopic pictures and opened up the field for moredetailed cell culture infection analyses Several immunofluorescence stainingprocedures have been developed, which can be applied after infection ofeukaryotic cell monolayers with pathogenic bacteria These proceduresenable a microscopic visualization and also a differential quantification ofadhered and internalized bacteria (Bergmann et al., 2009, 2013; Elm

in-et al., 2004; Jensch in-et al., 2010; L€uttge et al., 2012; Agarwal et al., 2013,

Table 1 Representative examples of bacterial pathogens analyzed in different cell culture models as highlighted in this review

Model system/cell type Bacterial strain References

Single cell type monolayer

Macrophages Helicobacter pylori Wang et al., 2009

Macrophages,

Dictyostelium

discoideum

Legionella pneumophila Steinert et al (1994), Allombert

et al (2014) , Steinert et al (2000) , Shevchuk and Steinert (2009) , and Skriwan

et al (2002)

D discoideum Mycobacterium marinum Meng et al (2014)

Dendritic cells Streptococcus pneumoniae Rosendahl et al (2013)

Epithelial and

endothelial cells

S pneumoniae Steinford et al (1989), Jensch

et al (2010), Bergmann et al (2009, 2013), L€uttge et al (2012), Agarwal et al (2013, 2014), Pracht et al (2005) , and Elm et al (2004)

Endothelial cells Streptococcus pyogenes Amelung et al (2011) and

Streptococcus agalactiae Nizet et al (1997)

Listeria monocytogenes Greiffenberg et al (1998)

Citrobacter feundii Badger et al (1999)

S pneumoniae Ring et al (1998) and Untucht

et al (2011)

Bloodecerebrospinal

fluid model N meningitides Steinmann et al (2013)

Lung coculture model Pseudomonas aeruginosa,

Klebsiella pneumoniae,

E coli

Hurley et al (2004)

3D culture model with scaffolds

Transwell system with

ECM

Chlamydia trachomatis Igietseme et al (1994), Kane and

Byrne (1998), Kane et al (1999) , and Dessus-Babus

2014; Pracht et al., 2005;Amelung et al., 2011;Nerlich et al., 2009; Rohdeand Chhatwal, 2013) In addition, fluorescence staining of the actin cyto-skeleton also visualized radical morphological changes of the eukaryoticcells, e.g., induced by streptococcal adherence (Bergmann et al., 2009).

Table 1 Representative examples of bacterial pathogens analyzed in different cell culture models as highlighted in this reviewdcont'd

Model system/cell type Bacterial strain References

L monocytogenes Cossart and Lecuit (1998)

Neisseria gonoorhoeae Hopper et al (2000)

S pyogenes Ochel et al (2014)

Epithelial airway tissue

model

S pneumoniae Fliegauf et al (2013)

P aeruginosa Woodworth et al (2008)

Rotary wall vessel culture

A549 lung epithelial

cells

Francisella tularensis David et al (2014)

P aeruginosa Carterson et al (2005)

E coli (EHEC/EPEC) Carvalho et al (2005)

Human skin equivalent Acinetobacter baumannii de Breij et al (2010) and Breij

et al (2014)

Tissue explants

Lung tissue explants Chlamydia pneumophila Rupp et al (2004)

L pneumophila J€ager et al (2014) and Shevchuk

et al (2014)

S pneumoniae Szymanski et al (2012)

Tonsil explants S pyogenes Bell et al (2012) and Abbot et al.

(2007)

Micro fluidic perfusion

HUVEC Staphylococcus aureus Pappelbaum et al (2013)

Complex responses of host immune systems are also studied using sion cultures with prepared and differentiated macrophages or other cellsderived from the lymphoid tissues.

suspen-This chapter will provide an overview about the broad spectrum of 2Dcell culture models in infection biology Beginning with the discussion ofkey aspects in using immortalized versus primary nonimmortalized eukaryoticmonolayers in infection models, the second part is focused on protozoa-basedmodels in infection biology Climbing to the next level of cell culturecomplexity, the advantages of coculture models generated by simultaneouscultivation of two or more different cell types will be described Thecocultivation technique is used to regenerate tissue barriers and will bedemonstrated by the examples of a transwell-based reconstruction of abloodebrain barrier (BBB) and a bloodecerebrospinal fluid barrier (BCSFB).2.1 Culture of Immortalized Cell Lines versus Primary CellCulture

Many valuable bacterial pathogenicity mechanisms have been elucidated ing 2D monolayer cell cultures indicating that certain scientific questions can

us-be answered and sometimes even require this kind of simplified cell culturetechnique A risk of using 2D monolayer cell culture models is the loss ordiminished expression of certain phenotypic characteristics that may mediatebacterialehost cell interactions This loss of phenotype results in progressivealterations in biochemistry, function, and morphology and increases withevery passage of the culture as the cells diverge from the source tissue pheno-type (Shaw, 1996) Of special importance are immortalized human cell lines,which have been used extensively for the study of hostepathogen interac-tions These stable cell lines combine the advantage of an indefinite life spanallowing passaging of several hundred times with low or moderate culturerequirements However, these lines exhibit aberrant properties attributable

to immortalization and artificial 2D growth conditions (Freshney, 2005).Thus, several studies targeting the investigation of pathogenecell inter-actions have been conducted with primary cells derived from different ani-mals such as primary porcine endothelial and epithelial cells (Vanier et al.,2007; Boekema et al., 2003) Nevertheless, in human-specific infectionbiology, the manual cell preparation of specific cell types from organ mate-rial remains a time-consuming and difficult process, retarded very often by alimited access to donor material, the lack of cell type-specific selectionmarker, a high risk of contamination, and suboptimal culture conditions.Fortunately, several companies such as PromoCell GmbH, Germany, and

PAA laboratories GmbH, Germany (part of GE Healthcare, USA) aremeanwhile specialized on eukaryotic cell culture techniques and provide

an enormously extended catalog of different cell types from human lung,skin, and gastrointestinal tract These catalogs also include more sensitivecell types such as primary epithelial, endothelial vascular cells, and smoothmuscle cells Most of the primary monotypic cells are characterized bycertain tissue-specific properties such as unique morphology, proliferationbehavior, and receptor expression pattern and are considered as appropriaterepresentatives of a specific tissue But this high level of specificity is based on

an expense of a higher sensibility toward culture conditions In general, mary cells require more complex cell culture media and a more experiencedhandling The cells are more susceptible toward any kind of stress and incontrast to immortalized cell lines, primary cells tend to rapidly lose theirhigh status of differentiation after being isolated from the native tissue(Freshney, 2005) Fulminant changes in cell differentiation and functionalspecificity are monitored especially for primary endothelial and epithelialcells, whereas immortalized cell lines tend to keep their main characteristicsfor longer proliferation rounds (Maqsood et al., 2013) Aflow cytometryanalysis of human primary lung endothelial cells, which has been performedexemplary for lung endothelial cells, clearly indicated significant changes inendothelial marker expression like platelet endothelial cell adhesion mole-cule 1 (PECAM1) within eight rounds of cell passaging and also depicted

pri-an inversion of the integrin expression pattern (Bergmpri-ann et al., unpublisheddata) This result is of highest interest, for example, with respect to the inves-tigation of receptor-mediated bacterial adherence and invasion in infectionbiology In order to ensure that the cells in culture express the same type andamount of surface receptors as in vivo, several cell culture infection analysesare performed additionally with primary cells directly prepared from therespective tissue In addition to the receptor profile, a further critical aspect

in long-term cell culture is the risk to lose the ability of cell type-specificfunctions For example, endothelial cells generate secretory WeibelePaladebodies, which arefilled with coagulation factors, vasodilation activators, andcytokines The functional study of these WeibelePalade bodies in infectionprocesses requires the use of primary cells in low passage, since the endothe-lial cells lose the ability of vesicle secretion after several in vitro passages(L€uttge et al., 2012) Notably, some cell types allow a specific manipulation

of a special cell activity, which may be required to answer a scientific tion A typical example is given by U937 cells, a human myelomonocyticcell line arrested at a monoblastic stage These cells have been extensively

used as a model to study cellular processes during different stages of cytic differentiation and are used as cell culture model for macrophage-mediated uptake of pathogens like Helicobacter, Mycobacteria, Legionella, andseveral more (Wang et al., 2009; Meng et al., 2014; Allombert et al.,2014; Steinert et al., 1994) The low basal level of phagocytic activity ofU937 can be increased by the induction of terminal monocytic differentia-tion through exposure to a combination of vitamin D3 and transforminggrowth factor-b1 (Wright et al., 1999) and allowed complex analyses of bac-terial uptake processes (Tacken and Batenburg, 2006) In general, immortal-ized cell lines remain basic components of standardized and easy-to-handlecell culture models But in order to circumvent the functional limitations,primary, differentiated, and functional active cells replace them in cell cul-ture models, as soon as specific cell functions and receptors interactionsare major part of scientific interest.

mono-2.2 Protozoa as Alternative Infection Models

Environmental protozoa are recognized as reservoirs and vehicles for severalimportant bacterial pathogens Thus, it is likely that on an evolutionary timescale, protozoaebacteria interactions have generated a pool of virulencetraits, which preadapted some bacterial species as human pathogens Theprotozoan mechanisms of phagocytosis, e.g., use signaling pathways andcytoskeleton proteins closely related to those of macrophages, neutrophils,and dendritic cells Moreover, it has been shown that many of the virulencefactors required for pathogenicity in mammals are also important for path-ogen survival during interactions with unicellular organisms such as Acantha-moeba castellanii or Hartmannella vermiformis (Hilbi et al., 2011; Steinert et al.,2003; Mody et al., 1993; Pearlman et al., 1988; Rowbotham, 1980) Afrequently utilized protozoan species, which allows highly relevant cross-species comparisons and mutant screenings on both sides of the hostepathogen interaction and the analyses of fundamental cellular processes, isDictyostelium discoideum (Shevchuk and Steinert, 2009; Steinert et al.,2003; Steinert et al., 2000; H€agele et al., 2000) The pathogens predomi-nantly analyzed in D discoideum are L pneumophila, Mycobacterium spp., Pseu-domonas aeruginosa, and Cryptococcus neoformans Unlike other protozoa, D.discoideum is amenable to genetic manipulation and combines many advan-tages such as easy cultivation, availability of cellular markers, the knowledge

of cell signaling pathways, and an elaborated molecular tool box ingly, the social amoeba D discoideum has become a prime model organism

Accord-in Accord-infection biology (Gerstenmaier et al., 2015; Weber et al., 2014; Bozzaro

et al., 2013; Clarke, 2010; Shevchuk and Steinert, 2009; Hilbi et al., 2007;Steinert et al., 2003; Skriwan et al., 2002; H€agele et al., 2000).

2.3 Coculture Infection Models

Primary and immortalized cell culture models are not perfect tives of the complex cellular environment found in organisms The devel-opment of several cocultivation techniques derived from the need togenerate tissue barriers for the simulation of pathogen-driven traversal dur-ing infection processes In coculture infection models, two or more singlecell types are cultivated simultaneously as monolayer, in suspension, or as amixed coculture, which typically utilizes adherent monolayers in combina-tion with a suspension cell type (Duell et al., 2011) An increasing amount

representa-of cocultures is generated as adjacent bilayers with the aim to create a morephysiological and complex surrounding such as a tissue barrier (Lindén

a porous membrane and the other cell type at the bottom This type ofcoculture is extensively used to elucidate the mechanism by which microor-ganisms like Streptococci, Helicobacter, and Staphylococci adhere, invade, andsignal to the host Moreover, these systems also enable an investigation ofimmune responses stimulated by the pathogens themselves or their secretedvirulence factors including determination of cytokine release and moni-toring of enhanced neutrophil transmigration (Lindén et al., 2007) Animportant general factor that can affect experimental design and the success-ful application of single culture as well as coculture models in infectionstudies is the effect of microbial activities on host cell viability Microbescan rapidly influence the viability of host cells through necrosis, apoptosis,and pyroptosis As a consequence, survival rates of the cocultured cells candecrease rapidly and alter the results (Wiegand et al., 2009) In addition to

a steady control of cell morphology and viability, negative side effects can

be reduced by decreasing infection time and bacterial load

In the following paragraphs, representative examples for coculture-basedinfection models will be described

2.3.1 Coculture-based generation of tissue barriers

The BBB and the BCSFB, separating brain interstitial space from blood, areanatomical and functionally unique barriers formed by the tight junctions

of brain microvascular endothelial cells (BMECs) (Pardridge, 1999;Rubinand Staddon, 1999) Therefore, the reconstruction of such complex barriersystems requires experienced knowledge about the detailed compositionand intercellular relationship between the involved cell types The proper-ties of the brain endothelium are supported and maintained by associatedcells, like astrocytes, pericytes, and microglia (Abbott, 2005) Astrocytesare known to induce and regulate many BBB characteristics and functions,namely the formation of tight junctions as well as the expression and asym-metrical localization of special enzymes, e.g., within transport systems(Abbott, 2005;Cecchelli et al., 2007) Viral, bacterial, fungal, and parasiticpathogens have been reported to breach the BBB and enter the central ner-vous system (CNS) through transcellular, paracellular, and the intracellular

“Trojan Horse” mechanisms Microscopic visualization techniques clearlydemonstrated transcellular invasion of BMEC by bacterial and fungal path-ogens including Escherichia coli (Huang et al., 1995, 2000; Kim, 2000),Group B Streptococcus (Nizet et al., 1997), Listeria monocytogenes (Greiffen-berg et al., 1998), Citrobacter freundii (Badger et al., 1999), S pneumoniae(Ring et al., 1998), Candida albicans (Jong et al., 2001), and several more.These studies revealed that the development of CNS inflammation isaccompanied by the release of cytokines (De Vries et al., 1997; Sun

et al., 1998) Moreover, barrier-forming coculture models identified crobial toxins and various microbial surface components as potent inducer

mi-of inflammation response mi-of the brain endothelium, which leads to eral damage of brain tissue and loss of barrier function (Stamatovic et al.,

collat-2008) As typical example for toxin-mediated damage of lung and brain sues, the pore-forming cytotoxin pneumolysin of S pneumoniae directlyactivates proinflammatory cellular cascades and induces apoptotic cell death(Hirst et al., 2004) Further destructive effects on tissue are also induced byactive cleavage of junctional proteins, which enables pathogen transmigra-tion of cellular barriers (Attali et al., 2008a) This kind of penetrationmechanism was reported for S pneumoniae bacteria recruiting the serineprotease activity of plasmin for degradation of catherines (Attali et al.,2008b) Paracellular penetration of the BBB has also been suggested forthe Lyme disease pathogen Borrelia burgdorferi (Garcia-Monco et al.,1990; Comstock and Thomas, 1991) and the syphilis-causing agent Trepo-nema pallidum (Haake and Lovett, 1994)

The complexity of BBB models ranges from transwell-based modelswith one or more cell types (astrocytes, pericytes, and endothelial cells)(Nakagawa et al., 2007) up to 3D model systems (Cucullo et al., 2007;Parkinson et al., 2003; Stanness et al., 1997) Thereby, the central compo-nent of most in vitro models of the BBB is BMEC, which exhibited a typicalcobblestone-like pattern at confluence, reconstitute in vitro their tight junc-tions and maintain specific cell properties up to passage 14 (Banks, 1999;Persidsky, 1999; Huang et al., 2000) BMEC-based in vitro models of theBBB have been used in various studies on cellular and molecular mecha-nisms of CNS infections caused by bacteria, virus, fungus, and parasites(Banks, 1999; Fusai et al., 2000; Huang et al., 2000; Jong et al., 2001).Initially, because of difficulties in isolating BMECs and growing them in cul-ture, most of the in vitro studies in pathogenesis of the CNS infection hadbeen carried out using large vessel endothelial cells such as human umbilicalvein endothelial cells (HUVEC) (Townsend and Scheld, 1995) Such kind

of systemic endothelial cells is still widely accepted and used as model cellsfor endothelial functions, although any kind of brain-specific property ismissing Thus, in disregard of species-specific interactions, bovine, murine,and human BMECs have been successfully used to dissect many pathogenicmechanisms of the CNS infection in vitro (Stins et al., 1997; Banks et al.,1998; Huang et al., 2000)

In order to study the complexity of the cell-to-cell interaction associatedwith the pathogenesis of bacterial pathogens such as Neisseria meningitidis, abilayer model with endothelial and epithelial cells (EC/EP bilayer) had beenestablished (Birkness et al., 1995) In this model system, endothelial andepithelial cells are separated by a microporous membrane, which allowsanalysis of bacterial transmigration through the multiple layers via micro-scopic visualization techniques Transwell-based transmigration modelsalso enable biological impedance determination by measurements of thetransepithelial electrical resistance (TEER), thereby monitoring the tightness

of the barrier and any loss of barrier function Interestingly, such coculturemodels revealed that many bacteria are efficiently penetrating from the api-cal surface through the epithelial layer and through the membrane to thebottom layer of endothelial cells without causing significant damage tothe host cells (Birkness et al., 1995)

A further typical bilayer cell model simulating lung tissue was developed

by Mul and colleagues and was composed of primary human endothelial(human papilloma virus-immortalized HUVEC cell line or primaryHUVECs) and lung epithelial cells (H292 or primary bronchial epithelial

cells) The cells were simultaneously cultured on opposite sides of transwellculture inserts (Mul et al., 2000) This model was applied for transmigrationanalysis of polymorphnuclear neutrophils (PMNs) (Parkos et al., 1991;Zeil-lemaker et al., 1996) PMN was labeled with calcein-AM prior to the start ofthe transmigration assay and after lysis, the amount offluorescence in each ofculture insert compartments or attached cells was measured in a spectroflu-orometer at the end of the experiment The above-mentioned examplesindicate the high level offlexibility of bilayer cell culture models to generateand maintain tissue barriers and to investigate specific cellular events corre-lating with the infection process at specific tissue sites.

2.3.2 Coculture of adherent cells and neutrophiles in suspensionCNS infection caused by microbial pathogens can lead to devastating neuro-logical disability and death Moreover, a critical point during the course ofcentral nervous system infection is the influx of leukocytes from the bloodinto the brain across the BBB but also across the BCSFB The pathogenshave developed a variety of strategies to breach the endothelium of theBBB or the choroid plexus epithelium of the BCSFB, which normally pre-vents entry of toxic substances into the brain interstitium For many impor-tant meningitis-causing pathogens such as Neisseria meningitides (Pron et al.,

1997), Haemophilus influenzae (Smith, 1987), E coli (Parkkinen et al., 1988),

L monocytogenes (Prats et al., 1992), Streptococcus suis (Sanford, 1987), andsome enteroviruses (Tabor-Godwin et al., 2010), experimental data suggestinvolvement of the choroid plexus during pathogen entry into the brain.Microbial infection of the CNS induces an increased transmigration rate

of polymorphnuclear neutrophils (PMNs) into the subcellular spaces asthefirst line of defense, promoted by an IL-8 release of epithelial or endo-thelial cells (Wittchen, 2009; Chin and Parkos, 2007) Various humanmodels of the BBB employing immortalized cell lines have been developed(Stins et al., 2001; Weksler et al., 2005; Muruganandam et al., 1997),whereas in vitro systems mimicking the BCSFB are limited to animal-derived cell material, including rat cell lines and primary porcine choroidplexus epithelial cells (PCPEC) (Haselbach et al., 2001; Shi and Zheng,2005; Gath et al., 1997; Kitazawa et al., 2001; Zheng and Zhao, 2002) Tis-sue barriers generated by PCPEC were used for leukocyte migration studiesand elucidated cell migration via the paracellular and the transcellular route(Wewer et al., 2011; Steinmann et al., 2013) The establishment of aninverted transwell filter system with porcine or human malignant choroidplexus papilloma cells (HIBCPP) comprising high barrier characteristics

enabled basolateral infection of host cells as well as investigation of gration of leukocytes from the pathophysiologically relevant blood side tothe apical cerebrospinal fluid side (Tenenbaum et a., 2013) This modelwas successfully applied to study the influence of N meningitidis infection

et al., 2013) Another infection model consisting of alveolar epithelial cells(A549) and human PMN grown on inverted culture inserts was used todetermine whether also bacteria such as P aeruginosa, Klebsiella pneumoniae,and E coli are capable of inducing PMN migration across these epithelialbarriers (Hurley et al., 2004) Calu-3 cells as well as primary human ATIIcells were also applied in inverted culture insert systems for similar objectives(Zemans et al., 2009) The afore-mentioned examples indicate the suitabilityand of transwell-based cell culture models for investigation of bacterial path-ogenicity mechanisms The very sophisticated requirements of reconstructedtissue barriers have been complied by several variations in supplementation

of scaffolding materials This kind of cell culture technique initialized thefield of 3D cell culture technique

3 3D CELL CULTURE

In many cases, the cocultivation of different cell types requires theassembly of a special kind of scaffold providing docking points and stimula-tion for adequate cell morphology, growth behavior, and survival Thesescaffolds are formed by nanometer-sizedfibers and pores, which are essential

to ensure a true 3D environment for the cell Studies spanning over twodecades of research provide several evidences that growing cells within3D scaffolds reduce the gap between cell cultures and physiological tissues.Thus, in order to maintain at most the physiological properties of cell tissues,great emphasis was laid on using 3D cell culture models that displayfunctional and phenotypic features of in vivo tissues (Fraley et al., 2010;Dhimolea et al., 2010) Meanwhile, an increasing amount of new technol-ogies, such as nanotechnology engineering, provides further evidence thatthe 3D in vitro cultivation of epithelial cells is crucial for such cells to senseand respond properly to receptor complex presentation (Discher et al., 2005;Nickerson et al., 2004) In fact, key events in the life cycle of a cell, such asproliferation, migration, and apoptosis, are regulated by organizing princi-ples that are determined by the cellular context (Bissel et al., 2002) Theseorganizing principles are maintained by cellecell and celleECM

(extracellular matrix) interactions, involve cytoskeletal orientation andsignaling, i.e., tyrosine phosphorylation and Rho/Ras/Rac activation, andestablish a 3D communication network that maintains the specificity andhomeostasis of the tissue (Kleinman et al., 2003) It is widely acceptedthat 3D cell cultures that reestablish such physiological cellecell and celleECM interactions can mimic the specificity of real tissues better than con-ventional 2D cultures 3D cultures are currently used in a broad range ofcell biology studies including tumor biology, cell adhesion, cell migration,and epithelial morphogenesis Example methodologies have also enteredthefield of infection biology and include the application of collagen-coatedpolycarbonate transwellfilter chamber inserts (Costar) and/or ECM-coatedfilter invasion chambers (BioCoat Matrigel; Becton Dickinson) to examinethe mechanisms of pathogen entry and intracellular fate of actin-recruitingListeria (Cossart and Lecuit, 1998), for studying interaction of Neisseria, Chla-mydia, and other pathogens with polarized and nonpolarized cell layers(Hopper et al., 2000; Igietseme et al., 1994; Kane and Byrne, 1998; Kane

et al., 1999; Dessus-Babus et al., 2002; Kazmierczak et al., 2001) and thereactivity to antibiotics and innate inflammatory response (Kenny et al.,

1997;Linzmeier and Ganz, 2005; Sansonetti, 2001)

3.1 Benefits and Limitations of 3D Scaffold

Tissue architecture is better represented by 3D cell culture than 2D cell ture Moreover, it has been reported that mechanical and biochemical cuesand cellecell communication are lost under the simplified and highly biasedconditions of 2D cell cultures This opinion is based on different cell cultureobservations, which will be discussed in the following paragraph Forexample,fibroblasts that migrate on a 2D substrate exhibit a different shapeand a different distribution of transmembrane adhesion proteins comparedwith fibroblasts within a 3D collagen scaffolding matrix (Walpita and Hay,2002; Cukierman et al., 2001; Meshel et al., 2005) Moreover, cells cultured

cul-in 3D reveal different gene expression levels compared with their 2D terparts Melanoma cells cultured onflat substrates upregulate and downregu-late other genes compared with melanoma cells cultured in 3D as spheroids Afurther example is given by the observation that mammary epithelial cells thatare cultured onflat plastic surfaces dramatically upregulate the expression ofmRNA that codes forb1-integrins By contrast, culturing on a recombinantbasement membrane induces expression levels of mRNA that are comparablewith those in the breast tissue (Delcommenne and Streuli, 1995) The impact

coun-of the mode coun-of cell cultivation system on expression coun-of cell surface receptors,

which are specifically required for bacterial interaction, is also highlighted in

an infection study with pathogenic Streptococci (Ochel et al., 2014) Here, agelatin-coated transwell system was used to analyze uptake of M1-proteinexpressing Streptococcus pyogenes bacteria by differentiated, polarized endothe-lial cells In addition to microscopic visualization of the bacterial uptake pro-cess, this cell culture infection model was also used to analyze the contribution

of the cytoskeleton and of the intracellular trafficking system to streptococcaluptake (Ochel et al., 2014) In addition to expression profiles of differentia-tion markers and tissue-specific receptors, it is often documented that thetopography and cytoskeletal organization of nonpolarized epithelial cellscultured in a 2D fashion in vitro on impermeable plastic or glass surfaces re-sembles that of simplefibroblasts In general, 2D cell cultures are characterized

by the following morphological properties: plasma membrane proteins areevenly distributed circumferentially; the perinuclear region is broad withthe Golgi complex juxtaposed to the nucleus; and microtubules extend hor-izontally out to the periphery in an orientation parallel to the basement mem-brane In contrast, epithelial cells cultured in vitro in a 3D polarizedorientation on suitable collagen or natural extracellular scaffold matrices arethree tofive times taller and have a different topography: plasma membraneproteins are separated into distinct apical and basolateral membranes by tightjunctions and junctional complexes, and are functionally compartmentalizedtherein; the Golgi complex assumes a supranuclear position in the apical cyto-plasm domain and microtubules are arranged vertically in an apical-to-basalaxis parallel to the lateral membrane (Guseva et al., 2007; Yeaman et al.,

1999) Biophysical experiments have highlighted further differences between2D and 3D cell migration: 2D culture situation allows only limited multishapemigration behavior, whereas, depending on the specific biological situation,cell migration can be of mesenchymal or amoeboid type, individual or collec-tive in clusters and multicellular sheets (Friedl, 2004) The migration speed ofcells also depends on the sterical and mechanical properties of the employed3D matrix (Zaman et al., 2006) Onflat surfaces, the speed with which cellsmigrate is related to the strength of the cell-surface adhesion, as determined byintegrin-dependent anchorage Whereas the maximum migration speed on2D substrates is reached in regimes of intermediate adhesiveness (DiMilla

et al., 1991) Many substances are used in the production of scaffolding materials including ceramics, synthetic polymers (polyurethanes, silicones,polyglycolic acid, polylactic acid, polyanhydrides, polyorthoesters), and natu-ral polymers (chitosan, glycosaminoglycans) and collagens Substantial varia-tions in matrix stability and lack in standardized reproducibility have to be

mentioned as current bottle necks, when using scaffold matrices in cell culturesystems This kind of difficulties can be partially overcome by using precoatedculture material provided by some companies During the last decades ofyears, collagen has become a favorable component of cell culture scaffolds.Several properties hallmark the suitability of collagens as cell culture embed-ding scaffold Depending on the collagen subtype, either robust proteinfibrils

or smaller net forming can be generated thereby providing a perfusable, inous hydrogel Collagen is a natural substrate for cells, and collagen gelsencourage cellular growth and have an impact on morphology, migration,and adhesion of cells (Kleinman et al., 1982) Nevertheless, because most ofthe currently available ECM gels are extracted from animals or cultured cells,quality control is difficult For example, the amount of undesired solublecomponents varies between batches, which reduces the reliability and repro-ducibility of the assay Progress is achieved with fully synthetic fibrousbiopolymer scaffolds, and gels of self-assembling synthetic oligopeptides arenow available for 3D cell cultures (for example, the commercially availablePuraMatrix) At pH and temperature conditions that are compatible withthat of tissue culture, the oligopeptide building blocks form a well-definedscaffold made of nanometer-sizedfibers These fibers and pores are essential

gelat-to ensure a true 3D environment for the cell (Gelain et al., 2006; Horii

et al., 2007) A further advantage is that such gels can be custom-tailoredwith specific amino acid sequences that are recognized by the cell’s adhesionreceptors (Zhang et al., 2004) With regard to infection biology, it has to bementioned that the choice of applied scaffold might affect the interaction be-tween pathogens and the tissue Some bacterial infections require the interac-tion with ECM proteins such as collagen by inducing an initial contact, which

is subsequently mediating adherence to cell surfaces On the other hand, specific interaction of bacteria with some scaffolding material might interruptany further interacting process with the tissue The following chapter willdiscuss typical examples for 3D scaffold-based cell culture models, whichare widely applied to analyze the interaction of pathogens with the humanand porcine BBB and the respiratory tract In addition, the beneficial effects

un-of microgravity on 3D cell culture will be discussed

3.1.1 Coculture-based reconstruction of BBB with matrix scaffoldThe BBB is a dynamic system and needs continuous induction processes, asevidenced by the fact that cells from brain microvessels can lose BBB fea-tures in monocultures (Reichel et al., 2003) EC/EP-bilayers mightimprove the cell layer complexity but several studies have shown that

astrocytes or related neuroepithelial cells contribute to the induction ofbarrier properties in BMEC (Kniesel and Wolburg, 2000) Based on thisobservation, an in vitro model consisting of both human BMEC and astro-cytes has been developed for analysis of penetration of HIV-1 and blood-derived monocytes into the CNS (Persidsky et al., 1997; Persidsky, 1999).Like the EC/EP bilayer system, human BMEC and astrocytes are cultured

in this model on opposite sides of a collagen-coated porous membrane intissue culture inserts This leads to a direct contact of the astrocyte end-feetnetwork with the BMEC monolayer This system has been successfullyused to study the mechanisms of virus-infected or activated leukocytemigration across the BBB (Persidsky, 1999) In another model, C6 astro-cytes were implemented as key element in an optimized in vitro BBBmodel (Hurst and Fritz, 1996; Untucht et al., 2011) C6 astrocytes are acell line derived from a rat glioma and secrete soluble factors inducingBBB-specific gene expression in brain endothelial cells (Hurst and Fritz,1996; Hurst et al., 1998; Kuchler-Bopp et al., 1999) In this model, humanbrain endothelial cells were embedded in a semisynthetic basement mem-brane (MatrigelÔ) The cocultivation with the C6 astrocytes allowed amore detailed characterization of BBB penetration mechanisms by try-panosomes and was approved to be suitable for analyses of drug penetrationproperties (Untucht et al., 2011; K€uhne et al., 2012)

3.1.2 Requirements of 3D tissue models generating aireliquid

surface

The commercially available epithelial airway tissue model (EpiAirwayTM)

of respiratory tract tissue enables cultivation on a microporous membrane

at an aireliquid surface It consists of normal human tracheal or bronchialepithelial cells and has been previously used for in vitro tests of nasalbioavailability (Agu et al., 2004; Chen et al., 2006) In order to confirmfunctionality of cells in culture and to circumvent the limitation of avail-ability of human tissue material for preparation of primary cells, animal tis-sue sources are used for generation of respiratory tract cell culture models.For this purpose, construction of an aireliquid interface culture of murinerespiratory epithelial was performed by preparation of respiratory epithelialcells from dissected mouse trachea The cells were cultured on collagen-coated transwell cell culture inserts and ciliogenesis was induced with aninsulin, transferrin, and selenous acid containing (ITS) premix, retinoicacid and by exposure of the apical cell surface to air (Fliegauf et al.,

2013) This kind of upper respiratory cell culture model effectively

simulates cell type-specific function including vigorous ciliary beating, lows long-term analyses and is widely used to study airway clearance mech-anisms of the host against respiratory pathogens like S pneumoniae and P.aeruginosa (Fliegauf et al., 2013; Woodworth et al., 2008) The EpiAir-wayTM has been shown to have a pseudostratified ciliated epithelium(Wengst and Reichl, 2010), which closely resembles the conditions inthe tracheobronchial epithelium (Wadell et al., 1999; Cotton et al., 1987).Using this principal model, a reconstruction of nasal mucosa was re-ported employing isolated human nasal fibroblasts in collagen matrixcovered by RPMI 2650 epithelial cells, which derived from squamouscell carcinoma of the nasal septum (Wengst and Reichl, 2010) These con-structs show a differentiated nonrespiratory-like epithelium and createpermeation barrier properties comparable to excised nasal mucosa (Wengstand Reichl, 2010) RPMI 2650 cells are able to form a confluent cell layerand develop sufficient TEER, as well as an appropriate permeation barrier,when cultured at the aireliquid interface (Bai et al., 2008) Under these cul-ture conditions, the cells also seemed to express tight junction proteins,although no pseudostratified or ciliated morphology as in vivo could beachieved Furthermore, this method has been shown to induce the differen-tiation of human nasal epithelial cells in primary and serial cultures betterthan in liquid-covered cultures (Lee et al., 2005; Yeh et al., 2007) A disad-vantage of EpiAirwayTM compared to an RPMI epithelial monolayermodel is the more complex handling of the constructs Nevertheless, thismodel is considered as a suitable permeation model for drug development,and as suitable infection model to study pathogenehost interaction,although no organotypic differentiation is achieved In general, the use of3D cell culture models enhances the repertoire of in vitro models to solvespecific scientific requirements, although they are not able to completely de-pict the whole in vivo situation.

al-3.2 MicrogravitydVariations of 3D Cell Culture Models

Studies conducted during spaceflight by the United States National tics and Space Agency showed that cell lines grown in suspension culture in lowshear microgravity tend to aggregate and exhibit morphologies more typical ofnative tissues (Unsworth and Lelkes, 1998; Hammond and Hammond, 2001).The low shear microgravity environment is also thought to be representative

Aeronau-of in vivo conditions in sites such as the uterus or within the brush bordermicrovilli, as predicted by mathematical modeling (Guo et al., 2000; Stockand Vacanti, 2001) Based on these findings, current attempts to improve

the phenotypic expression of eukaryotic immortalized host cells haveinvolved the use of low shear stress in 3D culture conditions (Barrila et al.,2010; Hjelm et al., 2010) To simulate low shear microgravity conditions

in the laboratory, rotating wall vessels (RWVs) were developed (Schwarz,R.P., Wolf, D.A Patent no 5.026.650, 1991) (Hammond and Hammond,

2001) These sterile cylindrical chambers are rotated on a specialized ized stand within an incubator at a velocity that allows the cells tofloat freely

motor-in suspension culture Gas exchange is provided through a gas-permeable icon rubber membrane ECM material such as collagen beads or sheets canalso be added to the bioreactor to support cell aggregation By growing cells

sil-in RWV bioreactors, tissue-derived cell cultures form 3D aggregate andexpress many of the in vivo phenotypic characteristics (Barrila et al., 2010;Carterson et al., 2005) Such an RWV-culture was generated with A549,

an immortalized human lung epithelial cell line, in order to study the tion process of Francisella tularensis (David et al., 2014) Data derived from un-infected probes indicate that monolayer-cultured A549 cells display a highlevel of cell-cycle activity and increased expression of oncogenes This expres-sion profile shows significant differences to normal lung epithelial cells.Conversely, the RWV-cultured A549 cells appeared to be differentiated,polarized, mucus producing cells, with a complex ECM, and are thereforebetter representatives of the in vivo lung epithelial cells (Carterson et al.,

infec-2005) A variation of this RWV-based 3D cell culture model is achieved

by providing microcarrier-bead surfaces for generation of polarized cellularmonolayers Such a model is established to study the intracellular infectionprocess of Chlamydia trachomatis (Guseva et al., 2007) Hereby, the in vitrocell culture system is based on cytodex collagen-coated microcarrier beads,rotating in a spinner bottle together with different cells such as McCoy(derived from knee joint synovial fluid from an arthritis patient), HEC-1B(endometrial carcinoma cells), or HeLa cells (derived from cervix carcinoma).The beads were kept slowly in suspension while a cell culture monolayergrows on their surface This system can be routinely monitored by phase mi-croscopy (Guseva et al., 2007) and has been confirmed as more suitable tostudy C trachomatis infection since higher infection rates were achieved inthis 3D system than in 2D in vitro cultures Nevertheless, it has to be kept

in mind that infection parameters also differ between various cell types used

in the 3D model system and are assumed to be a reflection of the tally different physiological functions between, for example, endometrialversus endocervical epithelial cells (Guseva et al., 2007) Moreover, compar-ison of 3D cell culture models using polarized cells with those using

nonpolarized cells revealed specific differences in the infection processdepending on the C trachomatis serovar subtype The luminal serovar E escapefrom polarized endometrial epithelial cells via their apical domains to spreadcanalicularly to the upper genital tract, whereas the invasive, lymphotrophicserovar L2 exit via the basal domains (Wyrick et al., 1989) It is assumedthat this latter result occurs due to a different microtubule orientation Inpolarized cells, microtubules are oriented parallel to the lateral membranewith an apical-to-basal axis, on which serovar L2 are known to traffic(Clausen et al., 1997) In another model system, human intestinal Int-407 cellsand A549 lung epithelial cells were grown in microgravity and have beenused as tissue models of infection with Salmonella enterica serovar Typhimu-rium and P aeruginosa infections, respectively (Nickerson et al., 2001;Carterson et al., 2005) In those studies, the tissue was propagated in micro-gravity, but the bacterial challenge was conducted in normal gravity It has to

be mentioned that microgravity is not only affecting eukaryotic host cells butalso alters gene expression of bacteria and also bacterial growth (Nickerson

et al., 2003, 2004) This is of importance since microarray analysis indicatedthat S enterica serovar Typhimurium grown in microgravity express differentsubsets of genes than those grown in a normal gravitationalfield (Wilson et al.,2002a,b)

4 ORGAN EQUIVALENTS AND TISSUE EXPLANTS

Regarding the cellular monolayer as simplest reductionist cell culturemodel, a higher complexity is achieved by coculture of different cell typesand 3D culture models using a scaffold material and rotating culture condi-tions The next level of complexity is reached by cell aggregates connected

in tissue-like manner called “cellular spheroids” or “organoids.” Cellularspheroids are simple 3D systems, which take advantage of the natural ten-dency of many cell types to aggregate without the requirement of externalscaffolding material The tissue counterparts of these suborgan structures arecommon to most of the epithelial organs and are known as“acini” in mam-mary tissue as well as lungs and “tubules” in the kidney (Carvalho et al.,

2005) These models offer several evaluation possibilities such as immunefluorescence microscopic visualization of bacteriaecell interaction Further-more, the organoids also allow histological analyses after formalin fixationand paraffin embedding And subsequent to trypsin-based cell dissociationfrom the organoids, single cell analysis is enabled by flow cytometry

Moreover, after homogenization of organoid tissues, enzyme activities such

as function of alkaline phosphatase can be determined using commerciallyavailable enzyme test kits (Carvalho et al., 2005) The simple spherical ge-ometry allows for relatively easy modeling of dynamic processes, such asgrowth and invasiveness of solid tumors (Stein et al., 2007; Jiang et al.,

2010) Cellular spheroids have become the system of choice for cally orientated biomedical studies (M€uller-Klieser, 1997; Sutherland, 1988;Sutherland et al., 1971), are applied in biotechnology (Kale et al., 2000), andare straightforward to apply in high-throughput screens (Ivascu and Kubbies,2006; Zhang et al., 2005) Spheroids can be obtained from single cultures orcocultures from a broad range of cell types (mono- or multicellular spher-oids) They are produced either by the hanging drop technique (Timmins

therapeuti-et al., 2005; Kelm therapeuti-et al., 2003) or by using RWV cultures or other cially available rotating cell culture systems (RCCS) (Synthecon, Houston,TX; Castaneda and Kinne, 2000; Unsworth and Lelkes, 1998; Nickerson

commer-et al., 2004) Thereby, the above-mentioned RWV-systems provide suitableconditions for the stable generation of tissue aggregates, such that growthcan be supported for several weeks The highest level of cell culturecomplexity is represented by reconstructed organoids and tissue equivalentsbased on special cell types cultured under organ or tissue-specific conditionsand by tissue explants The following examples are selected to elucidateadvantages and limitations by employing these highly sophisticated tissueculture systems in infection biology

4.1 Organoids and Tissue Equivalents Providing ComplexCell Systems“En Miniature”

The human ileocecal colorectal adenocarcinoma-HCT-8 cell line was selected

to establish a 3D intestinal organoid model for evaluation of the interactions ofprototypic strains of the enterohemorrhagic Escherichia coli (EHEC) andenteropathogenic Escherichia coli (EPEC) with the apical borders of such cellsunder conditions of microgravity (Carvalho et al., 2005) This model wasbased on hydrated cell culture sheets of small intestine submucosa (CookBiotech, West Lafayette, IN) providing a scaffold layer for HCT-8 cells.The HCT-8 cells derived from enterocytes at the junction of the large andsmall bowel (Tompkins et al., 1974) rapidly form 3D structures in a low shearmicrogravity culture generated by the RCCS (from Synthecon, Houston,TX) The size and surface area of the apical organoid cells appeared moretypical of normal intestinal epithelium than HCT-8 cells grown in mono-layers Similar to results obtained from simple 3D models, certain tissue

markers are better represented in 3D aggregates formed under tional conditions than in the same tissue types grown in 2D cultures(Nickerson et al., 2001; Carterson et al., 2005) This also includes E-cadherin,ZO-1, symplekin, and villin expression in HCT-8 organoids (Carvalho et al.,

microgravita-2005) In addition, expression of disaccharidases and alkaline phosphatase wassignificantly greater in the organoid-grown HCT-8 cells compared with cells

in 2D tissue culture From these observations, it has been concluded that theorganoid tissue was more differentiated and a better representative than, forexample, HCT-8 cells cultured in monolayers (Carvalho et al., 2005) Note-worthy, in contrast to polarized Caco-2 cells that grow in a single layer, theHCT-8 organoid cells formed multilayered structures As such, the interfacebetween the single layer of gut epithelial cells and the mesenchymal layer wasnot recapitulated in the organoid model Nonetheless, the morphology of theassembled cells varied according to their positions within the aggregate, withthe surface layer of cells most resembling normal gut epithelium Similarly,villin expression was concentrated at the tissue surface compared to thatseen in underlying cells This differential expression of villin is also seen invivo where epithelial cells express increasingly more villin as they matureand move to the tips of the villi (Maunoury et al., 1992)

It has been postulated that the cells at the surface of the organoid receivesignals for differentiation from the low-shear microgravity fluid environ-ment and thus provide a suitable model for bacterial interactions such asEHEC-infection at the lumenal surface of gut tissue (Carvalho et al.,

2005) As mentioned above, the microgravity also influences bacterialgene expression This was reported for Salmonella grown under micro-gravity, and also for the level of intimin production by EHEC (Carvalho

et al., 2005) A comparison of the data obtained from the 3D model with other infection models revealed some similarities such asintimin-mediated pedestal formation Nevertheless, results from other cellculture models also elucidated differences with regard to the adherencemechanisms depending on the expression of bundle forming pili and intimin(Carvalho et al., 2005; Donnenberg and Kaper, 1992; Hicks et al., 1998).Therefore, microgravity has an influence on bacteriaehost interaction,which has to be taken into consideration in comparison with data obtainedfrom different infection models

RCCW-In infection biology, several bacteriaehost interactions, such as ence, replication, and invasion, are strongly influenced by environmentalconditions, which in turn depends on the physicochemical barrier properties

adher-of the tissue surface and its nutrient availability For instance, in contrast to

the wet environment required for 3D models of the nasal epithelium, ciation of bacteria with the human skin is a process that takes place underrelatively dry conditions (Shepherd et al., 2009) Some tissue-engineeredair - exposed human skin models are 3D systems, which to a high degreemimic the native skin (El Ghalbzouri et al., 2008, 2009) Such epidermalskin equivalents are generated by culturing primary keratinocytes at theaireliquid interface on cell-free matrices (e.g., inert filters or de-epidermizeddermis) The keratinocytes proliferate, migrate, and differentiate duringepidermal development, resulting in skin equivalents that contain all layers

asso-of the native epidermis and elicit barrier properties with many similaritieswith the human skin (El Ghalbzouri et al., 2008; Thakoersing et al.,

2012) A 3D human skin equivalent has been applied to study skin zation by Acinetobacter strains and to evaluate the effects of disinfectants andother antimicrobial agents (Breij et al., 2014) Acinetobacter baumannii is able

coloni-to colonize the skin of hospitalized patients (Borer et al., 2007; Dijkshoorn

et al., 1987; Marchaim et al., 2007; Zeana et al., 2003), which can be asource of infection and spread to other patients and the environment Asexpected, the skin equivalent model revealed several differences compared

to other in vitro cell culture models using monolayer cell culture on plasticpetri dishes with respect to biofilm formation and induction of inflammatoryresponses (de Breij et al., 2010; Breij et al., 2014) and provided furtherevidence for the importance of the stratum corneum as a protective barrieragainst infections As a specific type of 3D cell culture model, tissue equiv-alents may provide a better environment to study pathogenehost interac-tions, although this latter example clearly indicates that the quality of themodel also depends on the presence of cell structures indispensable for thereplique of a specific in vivo situation

4.2 Tissue ExplantsdPiece of Reality

The ex vivo cocultivation of different cell types already represents a majorstep toward more system complexity and provides valuable data of thecell conglomerates in response to infective agents The cultivation of specifictissue explants grown in in vitro organ cultures (IVOC) provides a complexmulticellular and rather physiological environment for the study of hostepathogen interactions The benefits and limitations of tissue explants will

be discussed based on three examples: explants of human lung, human sils, and porcine skin

ton-Tissue explants play an important role in in vitro analysis of bacterial ogens showing restrictive species-specificities such as the human pathogen of

Legionnaires’ disease L pneumophila for the human lung Until recently,mammalian models such as guinea pigs, mice, rhesus monkeys, and marmo-sets were employed to address immunological, pathological, and pharmaco-logical questions (Baskerville et al., 1981; Blanchard et al., 1987; Fitzgeorge

et al., 1983), although data from these animal models cannot be easily alized because of important interspecies differences in the expression, func-tion, and localization of immune molecules (e.g., receptors, signalingintermediates, response molecules) Cell culture assays on the other handlack the complex interaction networks between the specialized cell typesand extracellular components in the human lung Therefore, tumor-free pul-monary tissue samples were developed as novel infection model for L pneu-mophila Histopathological analyses revealed that this approach narrows thegap between current infection models and actual human infections It allows

gener-to characterize tissue damage, bacterial dissemination, and the host’s lar response after an infection with L pneumophila including divers proteomeand transcriptome-based analyses (J€ager et al., 2014; Shevchuk et al., in press)

molecu-A similar ex vivo infection model of human lung tissue was also lished for the analysis of pneumonia infections induced by pneumococciwith the focus on detection of prostaglandin production in infected lung(Szymanski et al., 2012) This model was based on former description of aresolving infection model of mice lung tissue (Dockrell et al., 2003) and

estab-on an ex vivo model of acute chlamydial infectiestab-on, using human lung tissueexplants for interaction studies with the obligate intracellular pathogen Chla-mydia pneumophila (Rupp et al., 2004) The lung explant infection enabledthe determination of the prostaglandin expression profile, the cellular local-ization of prostaglandin production, and the decipherment of underlyingsignaling pathways (Szymanski et al., 2012)

As advantage of the lung explant model, the lung cell types are still nized into the unique lung architecture Therefore, cell-specific behaviorcan be studied Resident cells including alveolar macrophages are still pre-sent in the tissue and are capable of contributing to the observed innate im-mune response as shown previously byXu and coworker (2008) However,the lung explant infection does not allow for investigations of aspects of im-munity such as the recruitment of immune cells from the blood circulation(Szymanski et al., 2012)

orga-Tonsil explants represent a common model to mimic, for instance

S pyogenes-induced tonsillitis and pharyngitis, which are rarely ening, but among the most frequent of human infectious diseases affectingboth adults and children (Bell et al., 2012) With the aim to further

investigate bacterial adherence following the infection progress, ized human keratinocytes (HaCaT cells) were employed as a model forthe tonsil epithelium Additionally, tonsil explants of the control groupand of patients suffering from recurrent acute tonsillitis were challenged invitro with an M1-type S pyogenes (Abbot et al., 2007) A comparison be-tween the streptococcal effects on expression of host defense peptides in ton-sils derived from different patients revealed a marked variability relatedspecifically to individual tonsil samples within each of the groups In theseexperiments, this kind of variability tended to mask any significant trends.

immortal-A further application of tissue explants has been described in biofilmresearch While in vitro biofilm models using nonbiological surfaces suffi-ciently enable the generation of robust biofilms under certain physical andchemical environmental conditions, biological tissues such as dermal sub-strates contribute to a high degree to biofilm formation by providing attach-ment points and nutrition These tissues strongly influence size andconformation of the biofilm and alter the reactivity of the agents beingtested As example, porcine skin was applied in a microbial biofilm model

as both, the substrate for attachment and the primary source of nutrition(Yang et al., 2013) The produced biofilm resembled more closely the char-acteristics of biofilms that are found in human wounds In addition, thismodel can be used to assess the direct efficacy of antimicrobial dressingsagainst mature biofilms

In general, the use of tissue explants enables pathogenicity research in avery physiological environment ensuring species-specific unique character-istics Nevertheless, the increased complexity of the explanted tissue also ac-cumulates individual differences, which interfere with the cooperativeresponses of single cell types to the infection challenge In order to generateprincipal and reproducible statements, a defined selection of donors is animportant precaution for such type of infection model Unfortunately, theseculture systems are currently available only to facilities equipped to collectand rapidly convey biopsy materials to the research laboratory In addition,IVOC studies are limited to bacterialehost interactions that occur duringthe short life span of the organ biopsy material

4.3 Integration of Microfluidic Systems in 2D and 3D CellCulture

3D scaffolds provide tissue-like connectivity, although they are quitelimited in controlling the cell culture conditions, in nutrient and drug de-livery, and in running simultaneous assays during cell culture (Beebe,

2013) However, building 3D vascularized organs remains the maintechnological barrier to be overcome One of the major challenges is theinclusion of a vascular network to support cell viability in terms of nutrientsand oxygen perfusion Irrespective of cultivation in two or three dimen-sions, cell growth in vivo is significantly affected by the diffusion-limiteddistribution of oxygen, nutrients, and other molecules (Minchinton andTannock, 2006) Oxygen, nutrients, and other molecules are continuouslyconsumed and produced by cells Such dynamic distributions are notmimicked in conventional 2D or 3D cell culture (Martinez et al., 2008;Yamada and Cukierman, 2007) For an ideal 3D cell culture system,continuous nutrition and oxygen supply, and waste removal through theculture medium, must be ensured The microenvironment provided bymicrofluidic systems should be able to mimic this in vivo Thus, integration

of microfluidics with such 3D scaffolding systems allows dynamic ulation of culture conditions biochemically and biomechanically and pro-vides a microenvironment that allows formation of artificial tissues fromcultured cells (Trietsch et al., 2013; Choi et al., 2007) An exemplarytechnical setup, which combines tissue explant culture with microfluidicmedium supply, is depicted in Figure 2 In the 1990s, the development

manip-of microfluidic technology created a platform for highly complex anddynamic microenvironments that are controllable, reproducible, andadaptable to specific cell culture situations Microfluidic cell culture allowscontrollingfluid flow in the micrometer and nanometer scale in precisely

defined geometries and facilitates simultaneous manipulation and analysisstarting from a single cell level to larger populations and up to tissuescultured on fully integrated and automated chips (Mehling and Tay,

2014) Microfluidic systems add several benefits to in vitro cell culture.For example, the microscale dimensions of such microfluidic systems arecompatible with those of many microstructures and environments native

to in vivo systems For example, the distance between adjacent capillaries

in many in vivo animal tissue models is in the microscale region Moreover,some substrates like polydimethylsiloxane (PDMS) used in microfluidicdevices are permeable to oxygen, an important factor influencing cell pro-liferation A microfluidic system using PDMS chips also enables the culture

of blood vessel cells on the inner surface of microchannels whereflow andshear stresses of the blood circulation can be controlled (Fiddes et al., 2010).Thus, this system mimics functional aspects of the vasculature (Schimek

et al., 2013) Different technical solutions are commercially available andmost of them integrate multiple steps such as cell culture, cell sampling,

fluid control, cell capture, cell lysis, mixing, and detection on a singledevice The different microfluidic systems have been categorized asglass/silicon-based, polymer-based, and paper-based platforms, based onthe substrates used for microdevice fabrication (Li et al., 2012) A detailedoverview about the benefits of each single system is provided in a compre-hensive review by Li and colleagues (Li et al., 2012).

The use of microfluidic system provides new insights into infectionbiology For example, in order to study the interaction of pathogens withthe endothelial layer of blood vessels, human endothelial cells were effec-tively cultured to confluence in gel-free microslides, followed by infectionwith pathogenic Staphylococcus aureus bacteria (Pappelbaum et al., 2013).The application of a continuous and defined shear force induces significantchanges in cell morphology and replication rates resulting in cell shapestypical for the functional endothelium of blood vessels The polymer-based

Figure 2 A trendsetting tissue culture model combines the use of tissue explants with microfluidic perfusion systems (Photograph used with kind permission from ibidi GmbH, Germany.), thereby generating an autonomous ex vivo system This system acts as a vascular tubing, which supports the metabolism requirements and gas exchange of the tissue explant for longer cultivation periods.

microslides allowed a microscopic visualization of long vWF (von Willebrandfactor)-fibers, which were secreted by the endothelium in response to aninfection stimulus The vWF was bound by S aureus bacteria, therebydemonstrating the use of host-derived vWF as bacterial adhesion cofactor(Pappelbaum et al., 2013) The employment of microfluidic systems in gen-eration of synthetic capillaries also enabled investigation of pathogenicitymechanisms of the malaria causing agent Plasmodium falciparum Significantresults were obtained regarding the adhesion of infected red blood cells tohost cell ligands, the rheological responses to changing dimensions of capil-laries with shapes and sizes similar to small blood vessels, and the phagocytosis

of infected erythrocytes by macrophages (Antia et al., 2007) Moreover, amicrofluidic 3D bone tissue model was established for high-throughput eval-uation of wound-healing and infection-preventing biomaterials (Lee et al.,

2012) Osteoblasts are not able to form any 3D structure beyond a confluentlayer during conventional 2D culture However, 3D bone tissue-like struc-tures can be formed by long-term dynamic culture in microfluidic chambers

as a result of the proliferation of murine preosteoblasts, thereby forming aconfluent layer on the bottom chamber surface Based on a 3D aggregationwith produced collagen and calcium, the mineralized 3D tissue is formed bythe self-organization of osteoblasts in the microfluidic chambers The dy-namicflow of the microfluidic device emulates the nutrient and waste trans-port function of the microcirculation Moreover, the microscale geometricalform confines the culture chambers analogous to the size of micrometer-scalepores present in 3D scaffolds in tissue engineering The tissue morphology,which is generated by the microfluidic device, is composed of randomly ori-ented collagenfibers and contains calcified materials and osteocytes This tis-sue strongly resembles those of primary bone tissue and may provide a basismodel for further biochemical and mechanical processes involved in bacteria-induced bone infection (Lee et al., 2012)

Other examples are renal and hepatic cells that have also been cessfully cultured in close correspondence to the microarchitecture ofthe respective tissues (Lee et al., 2007) In addition to these homotypic tis-sue culture models, heterotypic tissue culture models that mimic therespective tissue closely both from a histologic as well as from a physiolog-ical and functional point of view have been achieved in microfluidic cellculture devices (Ho et al., 2013; Huh et al., 2010) This allows high-throughput pharmacological studies and might ultimately result in usingmicrofluidic cell culture systems also for regenerative purposes (Harink

suc-et al., 2013)

5 CONCLUDING REMARKS AND FUTURE

PERSPECTIVES

The use of eukaryotic cells or protozoa in 2D monolayer culture isirreplaceable in many areas of infection research Nevertheless, the physio-logical relevance of the information retrieved from in vitro studies is oftenquite limited and requires additional confirmation Dynamic distributions

of oxygen and nutrients are not mimicked in conventional 2D cell culture,and stimulatory effects of a highly complex 3D environment on cell growthand cell function are not recapitulated by the 2D cell culture In order toimprove the semiphysiological environment with single cell type culture,some trendsetting technical developments are currently in progress Withthe aim to analyze tissue-specific hostepathogen interactions, the currentdevelopment in monolayer-based 2D cell culture techniques is focused onthe preparation of highly specialized cell types from all relevant tissues,which maintain the high differentiation level in optimized cell culture me-dia This task is complemented by various projects aiming at optimization ofimmortalization techniques for the generation of cell lines with long-termreplication activities retaining cell type-specific differentiation and function-ality (Schmedt et al., 2012; Robin et al., 2015) The ability to cultivate moresensitive endothelial cells for longer periods of time in in vitro systems willoffer the possibility to expand the scientific knowledge about tissue destruc-tive inflammation responses like cytokine release, procoagulative reactions,and key signaling patterns In this regard, some recent developments in stemcell research might also invent new options to refine and facilitate the gen-eration of tissue models independent of the availability of specific cell mate-rial or tissues (McCracken et al., 2014) Primary keratinocytes have beensuccessfully used for generation of human epidermal equivalents (HEEs),but only a limited number of HEEs can be generated from one sample ofepidermis (G€otz et al., 2012) In order to develop an HEE model that can

be produced in an unlimited number of genetically identical units, humanembryonic stem cells (hESCs) were used to induce pluripotent stem cells(iPSCs) (Petrova et al., 2014) The stem cells are primary cells that arecapable of infinite proliferation and whose genetic footprint can be fullycharacterized The hESCs/iPSCs can be stimulated to differentiate into ker-atinocytes with gene expression profiles similar to those of normal humankeratinocytes (Petrova et al., 2014) These hESC/iPSC-derived keratino-cytes were used to generate HEEs in an aireliquid interface culture exposed

to a sequential high-to-low humidity environment HEEs generated from

hESC/iPSC-derived keratinocytes expose a functional permeability barriersimilar to native human skin and are indistinguishable from HEEs generatedfrom tissue-derived human keratinocytes under the same condition Thismodel has been considered as suitable alternative for elucidating the molec-ular mechanisms of barrier development, perturbation, and recovery, as well

as how mutations of genes involved in cornification and lipid metabolismaffect permeability barrier homeostasis (Mildner et al., 2010; Simpson

et al., 2010) The use of pluripotent stem cells for cell culture models mayprovide new options for generation of various tissue models in infectionbiology

The step from 2D cell culture toward 3D culture systems is hallmarked

by providing scaffolds for generation of tissue-like structures This higherlevel of complexity is represented by a huge and extending variety of tech-nical systems including coculture, microgravity, and various perfusablesetups The other side of the coin is that handling of 3D cell culture requires

a higher level of technical knowledge and experience in order to generatestable systems allowing reproducible infection analyses with respect to iden-tical matrix content and texture Organoids and tissue equivalents arecurrently the best in vivo-simulating systems, although the lack of a contin-uous nutrient supply and oxygen exchange limit possible applications.Recent reports describe successful attempts to combine engineered andexplanted tissues with a microfluidic device providing long-term perfusion.This trendsetting technique holds enormous potential for basic and appliedresearch in infection biology

ACKNOWLEDGMENTS

To provide a focused and clear review, we were forced to select representative examples and citations, and apologize for not mentioning all groups working with cell culture models in infection biology We thank Janine Rasch for providing photographs and the Deutsche For- schungsgemeinschaft (DFG) for financial support Authors’ own research was funded by DFG grants (BE 4570/4-1 and STE 838/8-1).

REFERENCES

Abbot, E.L., Smith, W.D., Siou, G.P., Chiriboga, C., Smith, R.J., Wilson, J.A., Hirst, B.H., Kehoe, M.A., 2007 Pili mediate specific adhesion of Streptococcus pyogenes to human tonsil and skin Cell Microbiol 9, 1822e1833.

Abbott, N.J., 2005 Dynamics of CNS barriers: evolution, differentiation, and modulation Cell Mol Neurobiol 25, 5e23.

Agarwal, V., Sroka, M., Fulde, M., Bergmann, S., Riesbeck, K., Blom, A.M., 2014 Binding

of Streptococcus pneumoniae endopeptidase O (PepO) to complement component C1q modulates the complement attack and promotes host cell adherence J Biol Chem.

289 (22), 15833e15844.

Agarwal, V., Kuchipudi, A., Fulde, M., Riesbeck, K., Bergmann, S., Bloom, A., 2013 tococcus pneumoniae endopeptidase O (PepO): a multifunctional plasminogen and fibro- nectin binding protein, facilitating evasion of innate immunity and invasion of host cells J Biol Chem 288 (10), 6849e6863.

Strep-Agu, R.U., Valiveti, S., Earles, D.C., Klausner, M., Hayden, P.J., Wermeling, D.P., Stinchcomb, A.L., 2004 Intranasal delivery of recombinant human parathyroid hor- mone [hPTH (1-34)], teriparatide in rats Endocr Res 30, 455e467.

Allombert, J., Lazzaroni, J.C., Bailo, N., Gilbert, C., Charpentier, X., Doublet, P., Vianney, A., 2014 Three antagonistic cyclic di-GMP-catabolizing enzymes promote differential Dot/Icm effector delivery and intracellular survival at the early steps of Legion- ella pneumophila infection Infect Immun 82, 1222e1233.

Amelung, S., Nerlich, A., Rohde, M., Spellerberg, B., Cole, J.N., Nizet, V., Chhatwal, G.S., Talay, S.R., 2011 The FbaB-type fibronectin-binding protein of Streptococcus pyogenes promotes specific invasion into endothelial cells Cell Microbiol.

13 (8), 1200e1211.

Antia, M., Herricks, T., Rathod, P.K., 2007 Microfluidic modeling of cell-cell interactions

in malaria pathogenesis PLoS Pathog 3, e99.

Attali, C., Durmort, C., Vernet, T., Di Guilmi, A.M., 2008a The interaction of Streptococcus pneumoniae with plasmin mediates transmigration across endothelial and epithelial mono- layers by intercellular junction cleavage Infect Immun 76, 5350e5356.

Attali, C., Frolet, C., Durmort, C., Offant, J., Vernet, T., Di Guilmi, A.M., 2008b coccus pneumoniae choline-binding protein E interaction with plasminogen/plasmin stim- ulates migration across the extracellular matrix Infect Immun 76, 466e476.

Strepto-Badger, J.L., Stins, M.F., Kim, K.S., 1999 Citrobacter freundii invades and replicates in human brain microvascular endothelial cells Infect Immun 67, 4208e4215.

Bai, S., Yang, T., Abbruscato, T.J., Ahsan, F., 2008 Evaluation of human nasal RPMI 2650 cells grown at an air-liquid interface as a model for nasal drug transport studies J Pharm Sci 97, 1165e1178.

Banks, W.A., 1999 Physiology and pathology of the bloodebrain barrier: implications for microbial pathogenesis, drug delivery and neurodegenerative disorders J Neurovirol.

5, 538e555.

Banks, W.A., Kastin, A.J., Arimura, A., 1998 Effect of spinal cord injury on the permeability

of the bloodebrain and blood-spinal cord barriers to the neurotropin PACAP Exp Neurol 151, 116e123.

Barrila, J., Radtke, A.L., Crabbe, A., Sarker, S.F., Herbst-Kralovetz, M.M., Ott, C.M., Nickerson, C.A., 2010 Organotypic 3D cell culture models: using the rotating wall vessel to study host-pathogen interactions Nat Rev Microbiol 8, 791e801 Baskerville, A., Fitzgeorge, R.B., Broster, M., Hambleton, P., Dennis, P.J., 1981 Experi- mental transmission of legionaires ’ disease by exposure to aerosols of Legionella pneumophila Lancet 2, 1389e1390.

Beebe, D.C., 2013 The use of cell lines to “model” ocular tissues: cautionary tales Invest Ophthalmol Vis Sci 54.

Bell, S., Howard, A., Wilson, J.A., Abbot, E.L., Smith, W.D., Townes, C.L., Hirst, B.H., Hall, J., 2012 Streptococcus pyogenes infection of tonsil explants is associated with a human beta-defensin 1 response from control but not recurrent acute tonsillitis patients Mol Oral Microbiol 27, 160e171.

Bergmann, S., Lang, A., Rohde, M., Rennemeier, C., Grashoff, C., Preissner, K.T., Hammerschmidt, S., 2009 Integrin-linked kinase is required for vitronectin-mediated internalization of Streptococcus pneumoniae by host cells J Cell Sci 122, 256e267 Bergmann, S., Schoenen, H., Hammerschmidt, S., 2013 The interaction between bacterial enolase and plasminogen promotes adherence of Streptococcus pneumoniae to epithelial and endothelial cells Int J Med Microbiol 303, 452e462.

Birkness, K.A., Swisher, B.L., White, E.H., Long, E.G., Ewing Jr., E.P., Quinn, F.D., 1995.

A tissue culture bilayer model to study the passage of Neisseria meningitidis Infect Immun.

63, 402e409.

Bissell, M.J., Radisky, D.C., Rizki, A., Weaver, V.M., Petersen, O.W., 2002 The organizing principle: microenvironmental in fluences in the normal and malignant breast Differen- tiation 70, 537e546.

Blanchard, D.K., Stewart 2nd, W.E., Klein, T.W., Friedman, H., Djeu, J.Y., 1987 Cytolytic activity of human peripheral blood leukocytes against Legionella pneumophila-infected monocytes: characterization of the effector cell and augmentation by interleukin 2.

J Immunol 139, 551e556.

Boekema, B.K., Stockhofe-Zurwieden, N., Smith, H.E., Kamp, E.M., van Putten, J.P., Verheijden, J.H., 2003 Adherence of Actinobacillus pleuropneumoniae to primary cultures

of porcine lung epithelial cells Vet Microbiol 93 (2), 133e144.

Borer, A., Gilad, J., Porat, N., Megrelesvilli, R., Saidel-Odes, L., Peled, N., Eskira, S., Schlaeffer, F., Almog, Y., 2007 Impact of 4% chlorhexidine whole-body washing on multidrug-resistant Acinetobacter baumannii skin colonisation among patients in a medical intensive care unit J Hosp Infect 67, 149e155.

Bozzaro, S., Buracco, S., Peracino, B., 2013 Iron metabolism and resistance to infection by invasive bacteria in the social amoeba Dictyostelium discoideum Front Cell Infect Micro- biol 3, 50.

Breij, E.C., de Goeij, B.E., Verploegen, S., Schuurhuis, D.H., Amirkhosravi, A., Francis, J., Miller, V.B., Houtkamp, M., Bleeker, W.K., Satijn, D., Parren, P.W., 2014 An anti- body-drug conjugate that targets tissue factor exhibits potent therapeutic activity against

a broad range of solid tumors Cancer Res 74, 1214e1226.

Carterson, A.J., Honer zu Bentrup, K., Ott, C.M., Clarke, M.S., Pierson, D.L., Vanderburg, C.R., Buchanan, K.L., Nickerson, C.A., Schurr, M.J., 2005 A549 lung epithelial cells grown as three-dimensional aggregates: alternative tissue culture model for Pseudomonas aeruginosa pathogenesis Infect Immun 73, 1129e1140.

Carvalho, H.M., Teel, L.D., Goping, G., O’Brien, A.D., 2005 A three-dimensional tissue culture model for the study of attach and efface lesion formation by enteropathogenic and enterohaemorrhagic Escherichia coli Cell Microbiol 7, 1771e1781.

Castaneda, F., Kinne, R.K., 2000 Short exposure to millimolar concentrations of ethanol induces apoptotic cell death in multicellular HepG2 spheroids J Cancer Res Clin Oncol 126, 305e310.

Cecchelli, R., Berezowski, V., Lundquist, S., Culot, M., Renftel, M., Dehouck, M.P., Fenart, L., 2007 Modelling of the bloodebrain barrier in drug discovery and development Nat Rev Drug Discov 6, 650e661.

Chen, S.C., Eiting, K., Cui, K., Leonard, A.K., Morris, D., Li, C.Y., Farber, K., Sileno, A.P., Houston Jr., M.E., Johnson, P.H., Quay, S.C., Costantino, H.R., 2006 Therapeutic utility of a novel tight junction modulating peptide for enhancing intranasal drug delivery J Pharm Sci 95, 1364e1371.

Chin, A.C., Parkos, C.A., 2007 Pathobiology of neutrophil transepithelial migration: cations in mediating epithelial injury Annu Rev Pathol 2, 111e143.

impli-Choi, N.W., Cabodi, M., Held, B., Gleghorn, J.P., Bonassar, L.J., Stroock, A.D., 2007 Microfluidic scaffolds for tissue engineering Nat Mater 6, 908e915.

Clarke, M., 2010 Recent insights into host-pathogen interactions from Dictyostelium Cell Microbiol 12, 283e291.

Clausen, J.D., Christiansen, G., Holst, H.U., Birkelund, S., 1997 Chlamydia trachomatis lizes the host cell microtubule network during early events of infection Mol Microbiol.

Cossart, P., Lecuit, M., 1998 Interactions of Listeria monocytogenes with mammalian cells ing entry and actin-based movement: bacterial factors, cellular ligands and signaling EMBO J 17, 3797e3806.

dur-Cotton, C.U., Stutts, M.J., Knowles, M.R., Gatzy, J.T., Boucher, R.C., 1987 Abnormal apical cell membrane in cystic fibrosis respiratory epithelium An in vitro electrophysio- logic analysis J Clin Invest 79, 80e85.

Cucullo, L., Hossain, M., Rapp, E., Manders, T., Marchi, N., Janigro, D., 2007 ment of a humanized in vitro bloodebrain barrier model to screen for brain penetration

Develop-of antiepileptic drugs Epilepsia 48, 505e516.

Cukierman, E., Pankov, R., Stevens, D.R., Yamada, K.M., 2001 Taking cell-matrix sions to the third dimension Science 294, 1708e1712.

adhe-David, J., Sayer, N.M., Sarkar-Tyson, M., 2014 The use of a three-dimensional cell culture model to investigate host-pathogen interactions of Francisella tularensis in human lung epithelial cells Microbes Infect 16, 735e745.

de Breij, A., Dijkshoorn, L., Lagendijk, E., van der Meer, J., Koster, A., Bloemberg, G., Wolterbeek, R., van den Broek, P., Nibbering, P., 2010 Do biofilm formation and in- teractions with human cells explain the clinical success of Acinetobacter baumannii? PLoS One 5, e10732.

De Vries, G., McLaughlin, A., Rhodes, J., 1997 The immunomodulatory actions of E-type prostaglandins Expert Opin Investig Drugs 6, 7e16.

Delcommenne, M., Streuli, C.H., 1995 Control of integrin expression by extracellular matrix J Biol Chem 270, 26794e26801.

Dessus-Babus, S., Darville, T.L., Cuozzo, F.P., Ferguson, K., Wyrick, P.B., 2002 ences in innate immune responses (in vitro) to HeLa cells infected with nondisseminating serovar E and disseminating serovar L2 of Chlamydia trachomatis Infect Immun 70, 3234e3248.

Differ-Dhimolea, E., Maffini, M.V., Soto, A.M., Sonnenschein, C., 2010 The role of collagen reorganization on mammary epithelial morphogenesis in a 3D culture model Biomate- rials 31, 3622e3630.

Dijkshoorn, L., Michel, M.F., Degener, J.E., 1987 Cell envelope protein profiles of tobacter calcoaceticus strains isolated in hospitals J Med Microbiol 23, 313e319 DiMilla, P.A., Barbee, K., Lauffenburger, D.A., 1991 Mathematical model for the effects of adhesion and mechanics on cell migration speed Biophys J 60, 15e37.

Acine-Discher, D.E., Janmey, P., Wang, Y.L., 2005 Tissue cells feel and respond to the stiffness of their substrate Science 310, 1139e1143.

Dockrell, D.H., Marriott, H.M., Prince, L.R., Ridger, V.C., Ince, P.G., Hellewell, P.G., Whyte, M.K., 2003 Alveolar macrophage apoptosis contributes to pneumococcal clear- ance in a resolving model of pulmonary infection J Immunol 171, 5380e5388 Donnenberg, M.S., Kaper, J.B., 1992 Enteropathogenic Escherichia coli Infect Immun 60, 3953e3961.

Duell, B.L., Cripps, A.W., Schembri, M.A., Ulett, G.C., 2011 Epithelial cell coculture models for studying infectious diseases: bene fits and limitations J Biomed Biotechnol.

2011, 852419.

El Ghalbzouri, A., Commandeur, S., Rietveld, M.H., Mulder, A.A., Willemze, R., 2009 Replacement of animal-derived collagen matrix by human fibroblast-derived dermal matrix for human skin equivalent products Biomaterials 30, 71e78.

El Ghalbzouri, A., Siamari, R., Willemze, R., Ponec, M., 2008 Leiden reconstructed human epidermal model as a tool for the evaluation of the skin corrosion and irritation potential according to the ECVAM guidelines Toxicol Vitro 22, 1311e1320.

Elm, C., Braathen, R., Bergmann, S., Frank, R., Vaerman, J.P., Kaetzel, C.S., Chhatwal, G.S., Johansen, F.E., Hammerschmidt, S., 2004 Ectodomains 3 and 4 of