key: cord-020764-5tq9cr7o
authors: Vertrees, Roger A.; Goodwin, Thomas; Jordan, Jeffrey M.; Zwischenberger, Joseph B.
title: Tissue Culture Models
date: 2010-05-21
journal: Molecular Pathology of Lung Diseases
DOI: 10.1007/978-0-387-72430-0_15
sha: 
doc_id: 20764
cord_uid: 5tq9cr7o

The use of tissue cultures as a research tool to investigate the pathophysiologic bases of diseases has become essential in the current age of molecular biomedical research. Although it will always be necessary to translate and validate the observations seen in vitro to the patient or animal, the ability to investigate the role(s) of individual variables free from confounders is paramount toward increasing our understanding of the physiology of the lung and the role of its cellular components in disease. Additionally, it is not feasible to conduct certain research in humans because of ethical constraints, yet investigators may still be interested in the physiologic response in human tissues; in vitro characterization of human tissue is an acceptable choice.

The use of tissue cultures as a research tool to investigate the pathophysiologic bases of diseases has become essential in the current age of molecular biomedical research. Although it will always be necessary to translate and validate the observations seen in vitro to the patient or animal, the ability to investigate the role(s) of individual variables free from confounders is paramount toward increasing our understanding of the physiology of the lung and the role of its cellular components in disease. Additionally, it is not feasible to conduct certain research in humans because of ethical constraints, yet investigators may still be interested in the physiologic response in human tissues; in vitro characterization of human tissue is an acceptable choice.

Tissue culture techniques have been utilized extensively to investigate questions pertaining to lung physiology and disease. The isolation and propagation of human bronchial epithelial cells has allowed investigators to begin to characterize the interactions and reactions that occur in response to various stimuli. Moreover, the culture of human airway smooth muscle has allowed researchers to investigate a pathologic cascade that occurs in asthma as well as other physiologic responses in the smooth muscle of the lung. Numerous lung cancer cell lines have been established to investigate their responses to chemotherapy and determine their biologic properties. Overall, the use of cultured human lung tissue has provided a windfall of information on the pathogenesis of diseases that affect the lung and on the basic physiology and development of the lung in general. Despite this wealth of information in the literature, this chapter is the fi rst to discuss the use of tissue culture models to examine the physiology and pathologic basis of lung diseases. In light of this, we briefl y discuss the history and principles behind the utilization of tissue culture. We then discuss the current use of tissue culture to examine many of the from Henrietta Lacks were cultivated into the fi rst immortal cell line-"HeLa." 5 HeLa cells are still one of the most widely used cell lines today. Since the 1950s, tissue culture has become fi rmly established as a mechanism to answer many questions in biomedical research. Today, tissue culture is widely used to investigate diseases that affect the lung, and through this work we have been able to increase our understanding of the pathologic cascades that occur in lung diseases, as well as the normal physiologies of the lung.

Tissue culture is a commonly used generic term for the in vitro cultivation of cells, attributed to the early cultures that generally consisted of heterogeneous cultures of crudely disaggregated tissues. Currently, many terms are used that can be encompassed by the term: organ culture, cell culture, primary explants, and ex vivo propagation all deal with the in vitro cultivation of cells or tissues. Cell culture in general can be applied either to primary cells (e.g., those with a fi nite life span) or to cell lines (e.g., HeLa cells). Additionally, these cultures can be either a homogenous or a heterogenous group of cells.

Primary cell culture involves the isolation of cells from a tissue by disaggregation. Single cell suspensions from tissues can be completed through either enzymatic digestion of extracellular matrix surrounding the cells-such as with ethylenediaminetetraacetic acid, trypsin, or collagenase-or mechanical disaggregation. These disaggregation procedures have the disadvantage of possibly injuring cells. If the cells of interest are adherent viable cells, they will be separated from nonviable cells when the medium is changed. Alternatively, viable cells can be separated from nonviable cells prior to culture by subjecting the single cell suspension to density gradient centrifugation (e.g., Hypaque). Primary cells have an advantage of possessing many of the biologic properties that they possessed in vivo because they are not transformed. Primary cells, unlike cell lines, are not immortal and have only a fi nite survival time in culture before becoming senescent. Variant cells, however, as well as those obtained from neoplastic tissue, may proliferate infi nitely, thus becoming immortal in vitro. This will eventually allow the immortal cell to take over the culture and can be thought of as a cell line. In general, primary human cultures will survive for 30-80 passages in vitro, although this number is dependent on cell type, conditions, and possibly other unknown factors. Primary cells are widely used to examine the effects of toxins, infectious agents, or other cellular interactions that would not be feasible in vivo. Primary cells have a disadvantage of being a heterogeneous mixture of cells upon primary isolation, with the type of cell obtained generally a component of the disag-gregation method used. The most common contaminant seen following isolation of primary cells is cells of mesenchymal origin (e.g., fi broblasts). However, advances have been made that allow the culture of homogenous populations of cells. For instance, cell surface molecules specifi c for the cells of interest may be tagged with monoclonal antibodies. Techniques such as fl uorescenceactivated cell sorting or the use of magnetic beads can be utilized to enrich the single cell suspension for the cell type of interest. Additionally, some investigators have recently exploited unique characteristics of certain cells, such as the presence of P-glycoprotein or multidrug resistance-associated proteins expressed on endothelial cells, to poison other contaminating cells in culture. 6 Another type of primary cell culture is "primary explants." This type of culture is not subjected to a disaggregation procedure like the primary cell technique described earlier. Therefore, single cell suspensions do not occur. Briefl y, tissue samples are dissected and fi nely minced. These tissue pieces are then placed onto the surface of a tissue culture plate. Following plating of tissue pieces, cells have been shown to migrate out of the tissue and onto the tissue culture surface. 7 This technique is useful when cells of interest may become damaged or lost in the disaggregation technique described earlier and is often used to culture human bronchial epithelial cells. 8 Cell lines are another useful source of cells to investigate questions in biomedical research. These cells have the advantage of being immortal as opposed to the fi nite life spans that primary cells possess. Additionally, they are generally well studied and characterized, leaving few experimental variables to worry about. These cells however, are prone to dedifferentiation-a process by which they lose the phenotypic characteristics of the cell from which they began. Many of the early cell lines were established from tumor tissue and as such possess abnormal growth characteristics. Newer cell lines have been established by molecular techniques such as inserting a telomerase gene into a cell to allow it to replicate infinitely. 9 Because of the phenotypic changes that allow cell lines to replicate infi nitely in culture, they are often a fi rst choice for experiments; however, they are also highly criticized in light of their nonnatural phenotype.

Organ culture, as the name implies, involves ex vivo culture of the whole or signifi cant portion of the organ. The main advantage to this type of culture is the retention and preservation of the original cell-cell interaction and extracellular architecture. This type of culture may be particularly important when experimental design necessitates the use of an ex vivo system, but researchers still need to retain the original organ architecture to answer questions posed. These types of cultures do not grow rapidly, however, and are therefore not suitable for experiments needing large numbers of a particular cell type. 10 Advantages and Limitations of Tissue Culture Tissue culture has become the penultimate tool of the reductionist biologist. The utilization of tissue culture as a research methodology has allowed investigators to study isolated interactions in its near-normal environment. These experiments by their very nature introduce artifacts; however, they do minimize the number of confounding variables that may affect a particular experiment. For instance, tissue culture allows investigators to determine the effects of one particular treatment on a particular cell type, which would not be feasible in vivo. Additionally, tissue culture models of disease allow investigators to obtain samples and make observations more readily than those done in vivo. However, it is the relative simplicity of experiments done in vitro that allows models of disease or physiology to come under frequent and warranted criticism. These models do not take into consideration the complexity of biologic systems. Diminishing possible confounding variables by culturing cells in vitro brings up the constant criticism of how applicable results are because of alterations of the normal cellular environment in vivo. For example, cell-cell interactions in vitro are reduced and unnatural. Moreover, the culture does not contain the normal heterogeneity and threedimensional architecture that is seen in vivo. This said, however, tissue culture biology has proved to be successful in many ways.

We have briefl y discussed the advantages that experimental systems using tissue culture affords researchers studying physiology and pathogenesis. Because of its ability to isolate individual variables and determine their role(s) in physiology, cell culture has become an integral tool in deciphering the pathologic cascades that occur in human disease. Diseases that affect lung are no exception.

Many diseases that affect the lung, and humans in general, are multifactorial. This begs the question how can cell culture, because of its reductionist nature only dealing with a minimal number of variables, help to solve the unknown questions and decipher the components involved in disease? Often, clinical observations, and the questions arising therein, have been the launching pad for investigation.

For instance, observations of massive infl ammation in the bronchoalveolar lavage samples of patients with acute respiratory disease syndrome (ARDS), consistent with damage seen in histologic samples, prompted investiga-tors to determine the role(s) of infl ammation in the etiology of ARDS. Through the use of cell culture, investigators were able to determine individual interactions that occurred in the disease process. Investigators have utilized culture models employing microcapillary endothelial cells under fl ow conditions to understand the role of proinfl ammatory cytokines in the cytokinesis and emigration of neutrophils in disease. Using a model of pulmonary endothelium under fl ow conditions allowed investigators to demonstrate the importance of certain proinfl ammatory cytokines in ARDS. 11 The role of inhaled toxicants in lung injury, and the mechanism(s) by which they cause disease, is another area of investigation that has utilized cell culture. Scientists have developed diverse and unique tissue culture systems that contain air-liquid barriers of lung epithelium and subjected these cells to various gaseous toxicants to determine what occurs following inhalation of various chemicals. Utilizing these types of systems, investigators are able to control the exposure time and other variables that may be diffi cult when determining inhaled toxicant effects in vivo. Moreover, the use of tissue culture, as opposed to an animal model, allows investigators to observe effects kinetically, without undue changes (e.g., sacrifi ce) and expense in the experimental model. 11 A tissue culture model also permits an investigator to observe multiple changes in real time, such as cellular integrity, cell signaling and intracellular traffi cking, protein expression changes, oxidant-induced cellular damage, and more. Deciphering each of these changes in an animal model would be extremely diffi cult; through employing a tissue culture model, researchers are able to tightly control the experimental system while isolating the events of interest. Further examples of how tissue culture models are currently being used to elucidate questions in lung physiology and disease are discussed later in the section on lung tissue cell lines.

Maintaining cells in vitro was initially a very diffi cult task. Many characteristics need to be fulfi lled before a successful cell culture occurs. Some of these characteristics are dependent on the type of tissue being studied; others may depend on specifi c requirements of the individual cells. Various chemically defi ned media are now available commercially to support the growth and differentiation of numerous cell types. The creation of defi ned media has allowed investigators to culture a multitude of cell types while controlling the local environment to answer pertinent questions. For example, glucose can be removed from a culture medium in order to study its effects on cellular metabolism, relative position in the cell cycle, and many other effects. Each chemical component is known in these media. Additionally, investigators can add growth factors to nourish their cell cultures.

The medium chosen when culturing cells in tissue culture must fi t two main requirements: (1) it must allow cells to continue to proliferate in vitro, and (2) it must allow the preservation of the certain specialized functions of interest. 7 The most common medium formulations used currently in lung research are Dulbecco's modifi ed Eagle's medium, minimum essential medium, RPMI 1640, and Ham's F-12. Occasionally, investigators develop new medium types to attain a formulation that optimizes their own experimental conditions. Fetal bovine serum is a common additive to most tissue culture media, although some investigators choose to forgo this additive for more defi ned supplementation. Additionally, others may choose sera from other sources such as human serum when culturing cells of human origin. Inactivation of complement by heat treating serum for 1 hr at 56°C was initially very popular in tissue culture. However, it has become clear that this treatment may in fact damage some of the proteinaceous growth factors present in the medium, rendering it less effective. Currently, many experts recommend heat inactivation only if the cell type of interest is particularly sensitive to complement. 12 More specifi c examples of medium utilized in lung tissue culture models are given later in the section on lung tissue cell lines.

When deciphering if the current culture conditions are suffi cient for the experimental design, the investigator must determine which cellular characteristics are important. Not only are the general characteristics, such as adhesion, multiplication, and immortalization of cell types important, but so are tissue-specifi c characteristics. Of importance to pulmonary research, the lung is a unique environment to simulate in vitro because of the air-liquid interface. Recently, investigators have made use of culture insert wells (e.g., Transwells, Corning) in order to study this interaction. 6 Cell Adhesion Nearly all normal or neoplastic human epithelial cells will attach with relative ease to tissue culture surfaces. Most tissue culture models utilizing tissue of lung origin fi t this description, with the notable exception of small cell lung carcinoma cell lines. However, for culture cells that may loosely adhere, or may not adhere at all, scientists coat tissue culture surfaces with extracellular matrix proteins. Incubating tissue culture surfaces with serum, as well as laminin, fi bronectin, or collagen, prior to culture has been shown to improve attachment of fi nicky cells. 8 These treatments also help in replicating the normal attachment of cells to extracellular matrix proteins in vivo.

The development of continuous cell lines may be serendipitous, as was the development of early cell lines. In brief, many investigators would continue splitting primary cell cultures until one or more cell clones became immortal. Unfortunately, the changes that generally occurred in culture led to cells with abnormal phenotypes that had undergone dedifferentiation. Today, many investigators choose to use molecular biology techniques, exploiting our current knowledge of oncogenic viruses and enzymatic processes of cellular aging to transform primary cells in vitro to an immortal phenotype. It is known that the large T antigen present in the SV (Simian virus) 40 virus is capable of transforming cells to an abnormal phenotype. 11, 13, 14 Moreover, transfection of primary cells with a transposase enzyme has also been shown to induce an immortal phenotypic change while preserving most normal cellular functions and phenotypes. 11 Dedifferentiation A commonly encountered problem in tissue culture is dedifferentiation. This loss of phenotype may be insignificant to the research at hand or it may be critical, and it must be dealt with on a case by case basis. When a cell culture undergoes dedifferentiation it is often unclear whether undifferentiated cells took over the culture of terminally differentiated cells or whether a primary cell of interest became immortal under the culture conditions.

The functional environment in which cells are cultured is critical when correlating experimental results to those seen in vivo. We previously alluded to the importance of the environment in which cells are cultured when discussing the advantages and limitations of tissue culture. Investigators have frequently strived to replicate integral in vivo environments in vitro in order to increase the signifi cance of their experimental results.

The development of cell culture insert wells (e.g., Transwells, Corning) has allowed investigators to culture bronchial or alveolar epithelial cells at an air-liquid interface. This ability allows investigators to begin to replicate a signifi cant aspect of these cells' functional environment in vitro, thereby increasing their understanding of the effects of gaseous particles on pulmonary epithelial cells. Alternatively, scientists have also cultured epithelial cells on a roller bottle apparatus. This method allows investigators to determine the amount of time the apical epithelial cell surface is in contact with the air.

Capillary cell cultures have also come under frequent criticism when cultured in a monolayer in a tissue culture plate. Investigators have been able to utilize gel matrices in which capillary cells form tubule-like structures, more closely replicating the architecture these cells maintain in vivo. Additionally, endothelial cells are constantly under fl ow conditions in vivo. Addressing this condition in vitro has allowed investigators to look at the role of endothelial cells during infl ammation-helping to increase the understanding of the role endothelium plays in acute lung injury.

At times, researchers may also choose to determine the effects of soluble factors (e.g., cytokines, hormones, neurotransmitters) from acute patients or animal models in a cell culture model. The milieu of soluble factors present in the serum that may play a role in a disease state is considerable. Moreover, these factors may have actions alone that are different when combined with other soluble factors. Reconstituting every factor presents a diffi culty in vitro and leaves the possibility that an unknown factor may be missing. To address this, investigators have harvested sera from patients or animal models and used these samples as additives in their media formulations. For instance, through the use of serum samples from an animal model of smoke/burn injury-induced acute lung injury, investigators have demonstrated that use of arteriovenous CO 2 removal in acute lung injury signifi cantly reduces apoptotic cell death in epithelial cells. 15 Lung Tissue Cell Lines: Establishment and Signifi cance

The diversity of research fi elds utilizing tissue culture models of lung diseases is extensive. In this section, we will give a brief overview of the main lung cell types that are being utilized in research today to answer pressing questions about lung physiology and the pathophysiology of pulmonary disease. Included in this discussion is also an overview of cell isolation and culture.

The use of normal human bronchial epithelial (HBE) cells is extensively reported in the literature. Based on a method pioneered by Lechner et al., 16 bronchial fragments obtained from surgery, autopsy, or biopsy specimens may be used as explants. The outgrowth of bronchial epithelial cells occurs readily from these explants when grown in medium supplemented with bovine pituitary extract and epidermal growth factor. Alternatively, these cells have also been demonstrated to grow in basal keratinocyte serum-free medium without supplementation; however, they demonstrate a slower growth rate and earlier senescence. 8 Cultures of HBE cells are valuable for determining the responses to toxic inhaled pollutants. In vitro exposure systems based on these methods have several advantages. First, in vitro exposure systems can be stringently controlled and reproduced much better than in animal systems; second, individual determination of the cell types' responses to pollutants allows for a better characterization of the individual involvement of the cell type to a biologic response. Finally, in vitro determination of the responses to toxic agents allows investigators to observe the reactions of human cells when testing in humans is not feasible because of ethical restraints.

In vitro study of the responses of bronchial cells to gaseous pollutants is not without its diffi culties. Wallaert et al. 17 have described these constraints well. Briefl y, because of the gaseous nature of the pollutants, culture systems should be designed that allow signifi cant exposure times to pollutants while also taking care to inhibit cells from drying out when exposed to air. To facilitate these experiments, roller bottle cultures have been developed that allow cells direct contact with the ambient air. Alternatively, cells have been grown on a membrane fi lter and cultured at an air-liquid interface, which allows constant exposure to the experimental treatment. The same type of experiments that are used to determine the responses of cells to inhaled toxicants have also been used to characterize responses to inhaled pharmaceuticals.

In addition to the characterization of responses to inhaled agents, epithelial cell cultures, notably alveolar epithelium obtained from fetal lung tissue, have allowed investigators to characterize the liquid transport phenotype that occurs in the developing lung. Characterization of the Cl − ion secretion system, which occurs in the distal lung epithelium throughout gestation, has been shown to be integral in the stimulation of growth of the developing lung by regulating liquid secretion. Likewise, a phenotypic switch of Na + absorptive capacity has been described toward the end of gestation, which is important for preparation of the lung to function postpartum and beyond. These culture systems have elucidated important physiologic changes that occur in the developing lung. Similar experiments have demonstrated that while ion transport plays a crucial role in this process other hormones and neurotransmitters are also important.

Pulmonary endothelial cells represent a unique type of endothelium because of their paradoxical responses to hypoxia. This uniqueness highlights the need to utilize cell culture models of pulmonary endothelium as opposed to other endothelia when interested in investigating their role(s) in pulmonary physiology. Several investigators have described the isolation and culture of pulmonary endothelial cells. Persistent pulmonary hypertension of the newborn, also known as neonatal pulmonary hyper-tension, is caused by a disorder of the pulmonary vasculature from fetal to neonatal circulation, culminating in hypoxemic respiratory failure and death. The inciting events that culminate in neonatal pulmonary hypertension are multifactorial. Despite this, decreased production of vasodilator molecules such as nitric oxide and prostaglandin I 2 in the pulmonary endothelium has been shown to be a critical component of disease progression. 18 

Primary cell cultures of human airway smooth muscle tissue can be obtained utilizing a method described by Halayko et al. 19 in which they isolated and characterized airway smooth muscle cells obtained from canine tracheal tissue. Briefl y, airway smooth muscle cells were obtained by fi nely mincing tissue and subjecting it to an enzymatic disaggregation solution containing collagenase, type IV elastase, and type XXVII Nagarse protease. Following generation of a single cell suspension, cells may be grown in Dulbecco's modifi ed Eagle's medium supplemented with 10% fetal bovine serum. Halayko et al. 20 obtained approximately 1.3 × 10 6 smooth muscle cells per gram of tissue using this method. Although Halayko et al. 21 pioneered this technique using trachealis tissue, many other investigators have obtained airway smooth muscle cells from a variety of biopsy specimens.

Airway smooth muscle hyperreactivity and hypertrophy has been known for nearly 100 years 2 to be an important end response of asthma. The use of airway smooth muscle in vitro has been vital toward delineating the pathologic steps that occur in asthma, as well as testing of potential therapeutics that may help to decrease the morbidity and mortality of asthma. Additionally, the relative paucity of in vivo models of asthma further illustrates the value of isolation and characterization of smooth muscle cells from asthmatic patients in vitro.

Using airway smooth muscle cell culture, investigators have characterized both the hypertrophic and hyperplastic growth of smooth muscle in individuals. Investigation of the potential stimuli that lead to airway smooth muscle proliferation and hypertrophy have led researchers to implicate the mitogen-activated protein kinase family members, extracellular signal-regulated kinase-1 and -2, and the phosphoinositol-3 kinase pathways in pathogenesis. 22 Additionally, mediators directing smooth muscle migration have also been observed in vitro and may play a role in the progression of asthma. Platelet-derived growth factor, fi broblast growth factor-2, and transforming growth factor-β (TGF-β) have all been shown to play a role in the migratory response of smooth muscle cells seen in asthma. 22 Additionally, contractile agonists such as leukotriene E 4 have been shown to potentiate the migratory responses seen with platelet-derived growth factor treatment. 22 Human airway smooth muscle cell culture has also been utilized to investigate possible pharmacologic interventions for the treatment of asthma. β 2 -Agonists have been shown to decrease the rate of DNA synthesis and likewise decrease the hyperplasia seen in airway smooth muscle cells in response to mitogenic stimuli through an increase in cyclic adenosine monophosphate. Like β 2agonists, glucocorticoids have similar antiproliferative activities.

Lung Cancer Tissue and the Development of Novel Therapeutics Culture of neoplastic cells from human tumors has allowed investigators to harvest a wealth of knowledge into the biology of lung cancers; moreover, these cultures have provided potential models to test potential therapeutics. The propagation of lung cancer cells in vitro has been covered in great depth previously. 8 In contrast to primary cell cultures, cultures of neoplastic cells are immortal, allowing their easy growth in culture with less chance of being overgrown by mesenchymal cells such as fi broblasts. The relative ease of growth in culture has led to many cell lines of lung cancer tissue. The National Cancer Institute, recognizing the need for a variety of lung cancer cell lines (both small cell and non-small cell), helped establish over 300 cell lines. 23 These lines are a wonderful resource for investigators given that they are extensively characterized, and many have full clinical data available. Moreover, many of these cell lines are now easily available through the American Type Culture Collection for a modest handling fee. Additionally, if investigators do not wish to use currently established lung cancer cell lines, obtaining clinical samples for use in tissue culture models is relatively easy. The same methods used to obtain biopsy specimens for clinical staging can also be used to begin cell cultures. Following culture and initial characterization of lung cancer cell lines, many investigators have demonstrated that lung cancer cell lines maintain a similar phenotype after establishment. Specifi cally, it has been verifi ed that injection of lung cancer cell lines into nude mice exhibit similar histopathology to the original tumor, indicating minimal change occurred following establishment of the cell lines.

Small cell lung carcinoma (SCLC) cell lines have been established from a multitude of biopsy specimens, including bone marrow, lymph nodes, and pleural effusions. 8, 24 Once viable cells have been obtained from clinical samples, cells are easily maintained in a basal cell culture medium such as RPMI 1640 in a humidifi ed incubator at 37°C and 5% CO 2 , although the initial isolations of SCLC lines utilized HITES and ACL-4 media. 25 Most established SCLC cell lines maintain a neuroendocrine phenotype in culture; however, Baillie-Johnson et al. 24 noticed considerable heterogeneity in the cell lines they established, highlighting the signifi cance that establishing a cell line from the clinical sample of interest may provide investigators with a line that possesses the exact phenotypic properties of interest.

Small cell carcinoma poses many diffi culties to surgical treatment, owing to its early and widespread metastasis. Therefore, combination chemotherapy is generally utilized in treatment. Unfortunately, despite initial sensitivities, SCLC tumors become resistant to further treatment. Utilizing in vitro cultures of SCLC cell lines, Sethi et al. 26 began to describe how extracellular matrix proteins can protect SCLC against apoptosis-inducing chemotherapeutics through β 1 -integrin-mediated survival signals. These data indicate that extracellular matrix proteins surrounding SCLC may play a role in the local recurrence seen in patients following chemotherapy in vivo and suggest novel therapeutics aimed at blocking these survival signals.

Non-small cell lung carcinoma (NSCLC) cell lines including squamous cell carcinoma, adenocarcinoma, and large cell carcinoma have all been established. Despite the fact that NSCLC cells comprise three distinct histologic cell types, all cell types can be established relatively easily. The primary treatment protocol for patients affl icted by NSCLC is generally surgical resection of the tumor; therefore, tumor cells for culture are readily available. These cell types can be grown under conditions similar to those described for SCLC.

Infectious diseases play a unique role in lung pathology in light of their roles as either important contributors or consequences of many lung diseases. For instance, certain lung diseases may predispose patients to infection: patients affl icted with obstructive lung diseases, as well as cystic fi brosis patients, commonly suffer from severe and recurrent bacterial infections. Additionally, patients may become superinfected following a viral respiratory infection. Systemic infections, such as gram-negative bacterial sepsis, may lead to lung diseases such as ARDS.

Human Type II Alveolar Pneumocytes and Acute Lung Injury/Acute Respiratory Distress Syndrome Pulmonary alveolar type II cells are a unique cell subset that carries out highly specialized functions that include synthesis and secretion of surfactant, a unique composition of lipoproteins that act to reduce surface tension at the alveolar air-liquid interface. 27 Defi ning the molecular mechanisms leading to production of surfactant by type II pneumocytes is important in many disease processes.

The pathogenic sequence that results in ARDS, the most severe manifestation of alveolar lung injury, is generally thought to be initiated by a systemic infl ammatory response. 28 Despite this knowledge, there still exist many questions about the initial triggers and pathologic steps that occur in ARDS. Greater understanding of these steps may help to develop new treatment regimes. Currently, treatment of ARDS consists of mechanical ventilation, which helps to stabilize blood gases. However, mechanical ventilation itself may provoke further infl ammation in the alveoli, thereby decreasing compliance and gas exchange in the alveoli. 29 The cell type of particular interest in ARDS and diffuse alveolar damage is the type II pneumocytes. [30] [31] [32] [33] [34] Until recently, studies trying to decipher the pathologic sequence in acute lung injury have had to rely on standard lung epithelial cell lines. Recently, however, human type II alveolar epithelial cells (pneumocytes) have been successfully isolated from fetal human lung tissue by collagenase digestion. 35 Briefl y, fetal lung tissues were minced and incubated in a serum-free medium containing dibutyryl cyclic adenosine monophosphate for 5 days. The tissue explants were then treated with collagenase and incubated with DEAE-dextran to eliminate contaminating fi broblasts. Cells were then plated onto tissue culture dishes treated with extracellular matrix derived from MDCK cells and cultured overnight in Waymouth's medium containing 10% serum. These steps resulted in relatively pure populations of human type II pneumocytes that were then cultured at an air-liquid interface. Using these methods, Alcorn et al. 35 were able to maintain a primary culture that retained the morphologic and biochemical characteristics of type II pneumocytes for up to 2 weeks.

Conventional Bioreactors and Three-Dimensionality: The Origins of Three-Dimensional Culture Carrel postulated that tissue development was linked to access to nutrient supply, noting that peripheral cells grew readily, and internal cells became necrotic presumably based on their distance from the nutrient source. To circumvent this issue, Carrel implemented cultures on silk veils, preventing the plasma clots of the growth media from deforming or becoming spherical, thus facilitating the internal cell's ability to obtain nutrient replenishment. Many attempts were made in standard culture systems (bioreactors) and other culture apparatuses to escape the constraints of two-dimensional cell culture, with the intent of yielding high-fi delity human and mammalian tissues, and thus emphasizing the need for development of three-dimensional biology.

Another famous researcher, Leighton, improved on Carrel's techniques in the 1950s and 1960s. Leighton's major contribution to three-dimensional culture technology was the introduction of the idea of a sponge matrix as a substrate on which to culture tissues. 36 , 37 Leighton fi rst experimented on cellulose sponges surrounded by plasma clots resident within glass tubes. He devised a system to grow 1-to 5-mm 3 tissue explants on sponges, using small amounts of chick plasma and embryo extract. After the mixture solidifi ed on the sponge Leighton added the nutrient media and inserted the "histoculture" in a roller apparatus to facilitate nutrient mass transfer. He experimented with many sponge combinations, discovering that collagen-impregnated cellulose sponges were optimal for sustaining the growth of native tissue architecture. 3, 38 Leighton was successful in growing many different tissue types on the sponge-matrix cultures. 3, 38 Leighton also found that C3HBA mouse mammary adenocarcinoma cells, when grown on sponge-matrix histoculture, aggregated "much like the original tumor, forming distinct structures within the tumors such as lumina and stromal elements, and glandular structures." An extremely important difference of this threedimensional histoculture from the standard two-dimensional culture is the apparent quiescence of the stromal component and the balanced growth of these cells with regard to the overall culture. Leighton further advanced the concept of three-dimensional histoculture to histophysiologic gradient cultures. 39 These cultures are conducted in chambers that allow metabolic exchange between "the pool of medium and the culture chamber by diffusion across a membrane." Histophysiologic gradient cultures mimic, to some degree, diffusion in tissues. 38 From the pioneering work of Carrel and Leighton, other methods of emulating three-dimensional cultures have been developed, such as embedding cells and tissues in collagenous gels of rat tail as per the techniques of Nandi and colleagues. Many of the advantages of threedimensional cultures seen by Leighton, Nandi, and others may be attributed to permitting the cells to retain their normal shape and special associations. 3 This global concept will be important as we begin to understand and recall the physical and environmental characteristics of the rotating-wall vessel systems.

Other methods of three-dimensional culture encompass a technique known as organ culture or culture on a fi lter, a strategy developed by Strangeways 40 and Fell and Robinson. 41 Tissue explants were grown on lens paper in a watch glass containing liquid culture medium. Browning and Trier 42 found "that for some tissues, it is critical to keep the cultures at the air-liquid interface," thus allowing the tissues to experience conditions similar to the in vivo environment.

Another strategy is the use of three-dimensional cultures known as proto-tissues, or aggregates of cells, used to form spheroids. This technique was popularized by Sutherland and colleagues more than 20 years ago when they manipulated aggregates of cells into a spherical confi guration by spinning agitation of the cells in spinner fl asks. 43 This technique produced pseudo-tissue-like organoids useful for research evaluations. Each of these methodologies will be of benefi t as we continue to examine strategies for achieving three-dimensional lung tissue constructs. 3, 38 Finally, membrane bioreactors are capable of retaining enzymes, organelles, and microbial, animal, and plant cells behind a membrane barrier, trapped in a matrix or adherent to the membrane surface. In 1963, Gallup and Gerhardt 44 fi rst used the membrane bioreactor for dialysis culture of Serratia marcescens. Immobilized enzyme microencapsulation was pioneered by Chang, 45 but Butterworth et al. 46 fi rst developed the enzyme membrane reactor to successfully accomplish starch hydrolysis with α-amylase. Likewise, for animal cell culturing, Knazek et al. 47 fi rst cultured human choriocarcinoma cells on compacted bundles of Amicon fi bers. Many reviews on the particular applications of hollow fi ber and immobilized bioreactant bioreactors for enzyme catalysts, microbial cells, and animal cell culture are available. [48] [49] [50] [51] [52] [53] As presented previously, tissue-engineering applications of three-dimensional function and structure are well known in medical science research. 54 In microgravity three-dimensional aggregates form, facilitating the expression of differentiated organotypic assemblies. Investigations to determine the effect of composite matrices, spiked with esterifi ed hyaluronic acid and gelatin, to augment osteochondral differentiation of cultured, bone marrow-derived mesenchymal progenitor cells and the effects of the matrix on cellular differentiation have been examined in vitro and in vivo. 54 Briefl y, empty and populated matrices cultured for 28 days, with and without TGF-β 1 demonstrated the following results. Cells implanted in the matrix produced a robust type II collagen extracellular matrix in vitro. Matrices placed in immunodefi cient mice yielded no differentiation in empty constructs, osteochondral differentiation in loaded implants, and an enhanced level of differentiation in preimplantation in vitro-cultured matrices containing TGF-β 1 . These results demonstrate the utility of three-dimensional matrix for presentation of bone mesenchymal progenitor cells in vivo for repair of cartilage and bone defects as well as indicate the efficacy for in vitro tissue engineering regimes. 54 These techniques lend themselves to microgravity and ground-based research tissue cultures alike.

Many earth-based laboratories are researching and developing hemopoietic bone marrow cultures of stem cell origin, and three-dimensional confi gurations are providing promising results as illustrated by Schoeters and coworkers. 55 They report that murine bone marrow cells, cultured under long-term hemopoietic conditions, produce mineralized tissue and bone matrix proteins in vitro but only when precipitated by the presence of adherent bone stroma cells in three-dimensional collagen matrices. At a concentration of 8 × l0 6 stromal cells, mineralization occurs in 6 days. In contrast, twodimensionally oriented marrow fragments at 1 × 10 7 cells require requires more than 10 days before mineralization can similarly be detected. 55 Two-dimensional long-term marrow culture facilitates and enhances expansion of the stromal component and rudimentary differentiation of osteogenic-like cells in the adherent stromal layer as verifi ed by type I collagen or cells positive for alkaline phosphatase. Production of osteonectin and osteocalcin, a bone-specifi c protein, combined with calcifi cation is observed only in threedimensional cultures. These studies demonstrate the need for and benefi t of three-dimensionality and the application to the microgravity environment. 55 As we can see, this further reinforces the quest for threedimensionality and the potential of modeling the microgravity environment.

Investigations clearly show the need for the application of three-dimensional study techniques in lung pathophysiologic studies. Interestingly, three-dimensional biology has facilitated full-scale investigations into most areas of tissue engineering, cell biology and physiology, immunology, and cancer research.

Anchorage-dependent cells are widely cultured on microcarriers. 56 Studies show that for the purposes of improved surface-to-volume ratio and scale up, the microcarrier suspension culture provides excellent potential for high-density cell growth. 57 In addition, microcarriers serve well as structural supports for three-dimensional assembly, the composite of which is the basis for threedimensional tissue growth. 58 Conventional culture systems for microcarrier cultures (i.e., bioreactors) use mechanical agitation to suspend microcarriers and thus induce impeller strikes as well as fl uid shear and turbulence at the boundary layer between the wall and the fl uid. Investigators have attempted to make a complete study of the most effi cient bioreactor designs and agitation regimens. 59 They concluded that virtually all stirred-tank bioreactors operate in the turbulent regimen. It has been demonstrated that bead-to-bead bridging of cells is enhanced signifi cantly at lower agitation rates in a stirred reactor. 60 Excessive agitation from either stirring or gas bubble sparging has been documented as a cause of cell damage in microcarrier cell cultures. 61, 62 To overcome the problems induced by these mechanisms, investigators developed alternative culture techniques such as porous microcarriers to entrap cells, 63 increased viscosity of culture medium, 64 bubble-free oxygenation, 65 and improved methods for quiescent inoculation. 66, 67 These steps decreased the damage attributed to turbulence and shear forces but failed to signifi cantly rectify the problems. Reactor systems of substantially increased volume exhibit less agitation-related cell damage. This is presumably because of the decreased frequency of cell-microcarrier contact with the agitation devices in the systems. Research-scale investigations do not afford the luxury of experimenting with large-scale production systems. Therefore, if a large-volume system is indeed more quiescent, an improved bioreactor system should emulate the fl uid dynamics present in the upper regions of large-scale reactors in which cells and microcarriers reside with minimal agitation. The problem, then, is to suspend microcarriers and cells without inducing turbulence or shear while providing adequate oxygenation and nutritional replenishment.

The term rotating-wall vessel comprises a family of vessels, batch fed and perfused, that embody the same fl uid dynamic operating principles. These principles are (1) solid body rotation about a horizontal axis that is characterized by (a) colocation of particles of different sedimentation rates, (b) extremely low fl uid shear stress and turbulence, and (c) three dimensional spatial freedom; and (2) oxygenation by active or passive diffusion to the exclusion of all but dissolved gasses from the reactor chamber, yielding a vessel devoid of gas bubbles and gas-fl uid interface (zero head space). 68, 69 Three-Dimensional Models of Lung Disease

Current cell culture models have shortcomings resulting in unreliable tumor growth, uncharacteristic tumor development, nonhuman tumors, and inadequate methods of detection. Cells propagated under traditional culture conditions differ widely in their expression of differentiated markers, adhesion receptors, and growth factor receptors compared with cells in situ or those grown as tissue-like structures. 70, 71 This is of concern because the phenotypic changes leading to malignant transformation often stem from alterations in the balanced and multifaceted roles of growth factors, receptors, and cytokines (reviewed by Herlyn et al. 71 ). With increasing evidence of the importance of adhesive contacts, paracrine cross-talk between different cell types, and signaling cascades that link the cell with a complex substratum, there is now recognition that models must be developed that better simulate these complexities. There is still much to learn about the dynamic relationships among the different phenotypes found in the normal lung and in lung cancers. Until a cell culture system is developed that allows differentiation to occur, 72 it is diffi cult to make any fi rm statement about relating effects in cell culture to clinical practice. Tissue engineering is very embryonic in development and currently nearly universally focused on building replacement tissues. A new technology developed at the NASA Johnson Space Center used to study colon cancer has been adapted to three-dimensional in vitro lung tissue culture models but has not been reported on to date.

Rotating-wall vessels are horizontally rotating cylindrical tissue culture vessels that provide controlled supplies of oxygen and nutrients with minimal turbulence and extremely low shear. 69 These vessels suspend cells and microcarriers homogeneously in a nutrient-rich environment, which allows the three-dimensional assembly of cells to tissue. Prior to seeding rotating-wall vessels (Synthecon, Inc, Houston, TX), cells were cultured in standard T fl asks (Corning, Corning, NY) in GTSF-2 medium (1993 PSEBM) in a humidifi ed 37°C, 5% CO 2 incubator. The rotating-wall vessels were seeded with 1-2 mg/mL Cultispher-GL microcarriers (Hyclone Laboratories, Inc., Logan, UT) followed by BEAS2-B or BZR-T33 cells (ATCC, Baltimore, MD) at a density of 2 × 10 5 cells/mL. Cultures were grown in the rotating-wall vessels for 14-21 days for formation of 3-to 5-mm diameter tumor masses. Rotating-wall vessel rotation was initiated at 25 rpm and increased as aggregate size became larger. Stationary control cultures were initiated under the same conditions using FEP Tefl on bags (American Fluoroseal, Columbia, MD). At 24-hour intervals pH, dissolved CO 2 , and dissolved O 2 were determined using a Corning 238 model clinical blood gas analyzer. Glucose concentration was determined using a Beckman 2 model clinical glucose analyzer (Beckman, Fullerton, CA). Cell samples were harvested every 48 hr and fi xed with Omnifi x (Xenetics, Tustin, CA) for immunohistochemistry or fi xed with 3% glutaraldehyde/2% paraformaldehyde in 0.1 M cacodylic buffer (Electron Microscopy Sciences, Fort Washington, PA) for scanning electron microscopy.

Cancer models already developed by NASA investigators include growth and differentiation of an ovarian tumor cell line, 72-74 growth of colon carcinoma lines, 72 and three-dimensional aggregate and microvillus formation in a human bladder carcinoma cell line. 74 In support as an appropriate model for cancer, even the most rudimentary three-dimensional cellular structures exhibit different phenotypes than cell lines cultured under twodimensional conditions. Properties such as responses to TGF-β, drug resistance to cisplatin or cyclophosphamide, and resistance to apoptosis are all altered in various types of cell aggregates. 75 Many investigations sustain consistent evidence that cells growing in three-dimensional arrays appear more resistant to cytotoxic chemoagents than cells in monolayer culture. 38 Li et al. found that spheroids were more resistant to cytosine arabinoside by 11-fold and methotrexate by 125-fold when compared with single cell suspensions. 76 Further monolayer cultures of colon carcinoma cells were sensitive to piericidin C in contrast to responses within in vivo colon tumors or three-dimensional slices of tumors grown in vitro. 77 Numerous other investigations have revealed increased levels of drug resistance of spheroids compared with single cell monolayers. 3, 38 Questions of poor diffusion and insuffi cient drug absorption within spheroids and a relatively frequent high proportion of resting cells have clouded differences in drug resistance, which could be the result of nutrient deprivation and hypoxia. Heppner and colleagues executed precise experiments that confi rmed threedimensional structure and function as the causative agent and was responsible for drug resistance rather than simple inaccessibility to nutrients or the drug concentration. Heppner embedded tumor specimens or cell aggregates in collagen gels, exposed the culture to various cytotoxic drugs, and compared the drug responses of the same cells in monolayers. These experiments revealed an increased resistance in the three-dimensional tumor arrays of a remarkable 1,000-fold greater than in monolayer cultures, and a similar result was seen in three-dimensional histocultures in collagen. The tumor cells grew in the presence of drug concentrations that rendered monolayers to a viability less than 0.1% of control cultures. Amazingly, Heppner observed that the cells became sensitive again when replated as monolayers and fi nally showed that even when exposed to melphalan and 5-fl uorouracil in monolayer cells transferred to collagen gels were again resistant based on three-dimensional architecture. Thus, the cells were exposed to the drugs as monolayers, facilitating access to the drugs, and, once the cells were transferred after drug exposure to a threedimensional structure, high resistance to the drugs was sustained. 38, [78] [79] [80] [81] Based on the caliber of data referenced above, Teicher et al. 82 serially passaged through multiple (10) transfers EMT-6 tumors in mice that were treated with thiotepa, cisplatin, and cyclophosphamide over a prolonged 6month period, thus producing extremely drug-resistant tumors in vivo. When these tumors were grown as monolayer cultures, they were as drug sensitive as the parental cells. Kobayashi and colleagues 83 grew the same in vivo drug-resistant tumor cell lines as spheroids in threedimensional arrays, and resistance was almost 5,000 times that of the parent line with selected drugs, an example being the active form of cyclophosphamide used in vitro. Similarly extreme resistance was also observed to cisplatin and thiotepa. This resistance was not seen in monolayer cultures, even when the monolayers were cultured on traditional extracellular matrix substrates. These experiments reconfi rmed that cells in a three-dimensional array are more drug resistant than monolayer cells in vitro and demonstrated that three-dimensional cellular confi gurations can and do become resistant to super pharmacologic doses of drugs by forming compact structures. 38 Rotating-Wall Vessel Tumor Models Several important human tumor models have been created in rotating-wall vessel cultures, specifi cally, lung, prostate, colon, and ovarian. 14, 58, 73, 84 Many of these models involve cancers that are leading killers in our society. We present two such examples in this section, colon and prostate carcinoma. As previously reviewed, the literature indicates the remarkable difference between chemotherapeutic cytotoxicity in two-dimensional and three-dimensional cellular constructs, which may be predicated on a number of criteria. Therefore, a threedimensional tumor model that emulates differentiated in vivo-like characteristics would provide unique insights into tumor biology.

Goodwin et al. 58 detail the fi rst construction of a complex three-dimensional ex vivo tumor in rotatingwall vessel culture composed of a normal mesenchymal base layer (as would be seen in vivo) and either of two established human colon adenocarcinoma cell lines, HT-29, an undifferentiated line, and HT-29KM a stable, moderately differentiated subline of HT-29. Each of these engineered tumor tissues produced tissue-like aggregates (TLAs) with glandular structures, apical and internal glandular microvilli, tight intercellular junctions, desmosomes, cellular polarity, sinusoid development, internalized mucin, and structural organization akin to normal colon crypt development. Necrosis was minimal throughout the tissue masses up to 60 days of culture while achieving >1.0 cm in diameter. Other notable results included enhanced growth of neoplastic colonic epithelium in the presence of mixed normal human colonic mesenchyme. These results mimic the cellular differentiation seen in vivo and are similar to results obtained with other tumor types.

Prostate carcinoma has also been modeled in the rotating-wall vessel system by several investigators. [85] [86] [87] One of the most comprehensive descriptions of these engineered tissues is detailed by Wang et al. 88 In that review, the authors describe the ability of the rotatingwall vessel system to recapitulate human prostate carcinoma (LNCaP) and bone stroma (MG63) to illuminate the evolution of prostate tumorigenesis to the metastatic condition. In particular, the LNCaP and ARCaP models represented in the review are known to be lethal in the human, being androgen independent and metastatic. Rotating-wall vessel TLA engineering also allowed indepth study of epithelial and stromal interactions, which are the facilitating elements of the continuance of LNCaP prostate-specifi c antigen production in vitro. When LNCaP was cultured in three dimensions without stroma, production of prostate-specifi c antigen ceased and metastatic markers were not observed. The authors outline the process of malignant transformation, demonstrating that these metastatic models are only possible in threedimensional TLAs and are achieved by specifi c geometric relationships in three-dimensional confi guration. Furthermore, they show through direct comparison with other culture systems the advantages of the rotating-wall vessel system to allow synergistic relationships to study this disease state. 88 Unlike two-dimensional models, these rotating-wall vessel tumor tissues were devoid of metabolic and nutrient defi ciencies and demonstrated in vivo-like architecture. These data suggest that the rotating-wall vessel affords a new model for investigation and isolation of growth, regulatory, and structural processes within neoplastic and normal tissues.

In this section, we explore the utility of rotating-wall vessel TLAs as targets for microbial infection and disease. Several studies have been conducted recently that indicate that three-dimensional tissues respond to infective agents with greater fi delity and with a more in vivo-like response than traditional two-dimensional cultures. Nickerson et al. 89 describe the development of a threedimensional TLA engineered from INT-407 cells of the human small intestine, which were used as targets for the study of Salmonella typhimurium. In this study, threedimensional TLAs were used to study the attachment, invasion, and infectivity of Salmonella into human intestinal epithelium. Immunocytochemical characterization and scanning and transmission electron microscopic analyses of the three-dimensional TLAs revealed that the TLAs more accurately modeled human in vivo differentiated tissues than did two-dimensional cultures. The level of differentiation in the INT-407 TLAs was analogous to that found in previously discussed small intestine TLAs 72 and from other organ tissues reconstructed in rotatingwall vessels. Analysis of the infectivity studies revealed Salmonella attached and infected in a manner significantly different from that in control two-dimensional cultures. During an identical exposure period of infection with Salmonella, the three-dimensional TLAs displayed a minor loss of structural integrity when compared with the two-dimensional INT-407 cultures. Furthermore, Salmonella demonstrated a greatly reduced ability to adhere, invade, and induce the apoptotic event in these INT-407 three-dimensional TLAs than in twodimensional cultures. This result is not unlike the in vivo human response. Two-dimensional cultures were significantly damaged within several hours of contact with the bacteria; conversely, although "pot marks" could be seen on the surfaces of the three-dimensional TLAs, they remained structurally sound.

Cytokine analysis and expression postinfection of three-dimensional TLAs and two-dimensional cultures with Salmonella exhibited remarkable differences in expressed levels of interleukin (IL)-1α, IL-1β, IL-6, IL-1Ra, and tumor necrosis factor-α mRNAs. Additionally, noninfected three-dimensional TLAs constitutively demonstrated elevated levels of TGF-β 1 mRNA and prostaglandin E 2 compared with noninfected two-dimensional cultures of INT-407. 89 As previously stated, traditional two-dimensional cell monolayers lack adequate fi delity to emulate the infection dynamics of in vivo microbial adhesion and invasion. The respiratory epithelium is of critical importance in protecting humans from disease. Exposed to the environment, the respiratory epithelium acts as a barrier to invading microbes present in the air, defending the host through a multilayered complex system. 90 The three major layers of the human respiratory epithelium are pseudostratifi ed epithelial cells, a basement membrane, and underlying mesenchymal cells. Ciliated, secretory, and basal epithelial cells are connected by intercellular junctions and anchored to the basement membrane through desmosomal interactions. Together with tight junctions and the mucociliary layer, the basement membrane maintains the polarity of the epithelium and provides a physical barrier between the mesenchymal layer and the airway. 91, 92 Infi ltrating infl ammatory and immune cells move freely between the epithelial and subepithelial compartments.

Airway epithelial cells play a vital role in host defense 90 by blocking paracellular permeability and modulating airway function through cellular interactions. Ciliated epithelial cells block invasion of countless inhaled microorganisms by transporting them away from the airways. 93 As regulators of the innate immune response, epithelial cells induce potent immunomodulatory and infl ammatory mediators such as cytokines and chemokines that recruit phagocytic and infl ammatory cells that remove microbes and enhance protection. 90, 91, 94, 95 Ideally, cell-based models should reproduce the structural organization, multicellular complexity, differentiation state, and function of the human respiratory epithelium. Immortalized human epithelial cell lines, such as BEAS-2B, 96 primary normal human bronchial epithelial cells, 97 and air-liquid interface cultures, 98 are used to study respiratory virus infections in vitro. Traditional monolayer cultures (two-dimensional) of immortalized human bronchoepithelial cells represent homogenous lineages. Although growing cells in monolayers is convenient and proliferation rates are high, such models lack the morphology and cell-cell and cell-matrix interactions characteristic of human respiratory epithelia. Thus, their state of differentiation and intracellular signaling pathways most likely differ from those of epithelial cells in vivo. Primary cell lines of human bronchoepithelial cells provide a differentiated model similar to the structure and function of epithelial cells in vivo; however, this state is short lived in vitro. 97, 99 Air-liquid interface cultures of primary human bronchoepithelial cells (or submerged cultures of human adenoid epithelial cells) are grown on collagen-coated fi lters in wells on top of a permeable fi lter. These cells receive nutrients basolaterally, and their apical side is exposed to humidifi ed air. The result is a culture of well-differentiated heterogeneous (ciliated, secretory, basal) epithelial cells essentially identical to airway epithelium in situ. 98, 100 Although this model shows fi delity to the human respiratory epithelium in structure and function, maintenance of consistent cultures is not only diffi cult and time consuming but also limited to small-scale production and thus limits industrial research capability.

True cellular differentiation involves sustained complex cellular interactions [101] [102] [103] in which cell membrane junctions, extracellular matrices (e.g., basement membrane and ground substances), and soluble signals (endocrine, autocrine, and paracrine) play important roles. [104] [105] [106] [107] This process is also infl uenced by the spatial relationships of cells to each other. Each epithelial cell has three membrane surfaces: a free apical surface, a lateral surface that connects neighboring cells, and a basal surface that interacts with mesenchymal cells. 108 Recently viral studies by Goodwin et al. 109 and Suderman et al. 110 were conducted with rotating-well vessel-engineered TLA models of normal human lung. This model is composed of a coculture of in vitro threedimensional human bronchoepithelial TLAs engineered using a rotating-wall vessel to mimic the characteristics of in vivo tissue and to provide a tool to study human respiratory viruses and host-pathogen cell interactions. The TLAs were bioengineered onto collagen-coated cyclodextran beads using primary human mesenchymal bronchial-tracheal cells as the foundation matrix and an adult human bronchial epithelial immortalized cell line (BEAS-2B) as the overlying component. The resulting TLAs share signifi cant characteristics with in vivo human respiratory epithelium, including polarization, tight junctions, desmosomes, and microvilli. The presence of tissuelike differentiation markers, including villin, keratins, and specifi c lung epithelium markers, as well as the production of tissue mucin, further confi rm these TLAs differentiated into tissues functionally similar to in vivo tissues. Increasing virus titers for human respiratory syncytial virus (wtRSVA2) and parainfl uenza virus type 3 (wtPIV3 JS) and the detection of membrane-bound glycoproteins (F and G) over time confi rm productive infections with both viruses. Viral growth kinetics up to day 21 pi with wtRSVA2 and wtPIV3 JS were as follows: wtPIV3 JS replicated more effi ciently than wtRSVA2 in TLAs. Peak replication was on day 7 for wtPIV3 JS (approximately 7 log 10 particle forming units [pfu] per milliliter) and on day 10 for wtRSVA2 (approximately 6 log 10 pfu/mL). Viral proliferation remained high through day 21 when the experiments were terminated. Viral titers for severe acute respiratory syndrome-coronavirus were approximately 2 log 10 pfu/mL at 2 day pi.

Human lung TLAs mimic aspects of the human respiratory epithelium well and provide a unique opportunity to study the host-pathogen interaction of respiratory viruses and their primary human target tissue independent of the host's immune system, as there can be no secondary response without the necessary immune cells. These rotating-wall vessel-engineered tissues represent a valuable tool in the quest to develop models that allow analysis and investigation of cancers and infectious disease in models engineered with human cells alone.

We have explored the creation of three-dimensional TLAs for normal and neoplastic studies and fi nally as targets for microbial infections. Perhaps Carrel and Leighton would be fascinated to know that from their early experiments in three-dimensional modeling and the contributions they made has sprung the inventive spirit to discover a truly space age method for cellular recapitulation.

The Biochemical Basis of Pulmonary Function

The pathology of bronchial asthma

Three-dimensional histoculture: origins and applications in cancer research

Histoculture of human breast cancers

Studies on the propagation in vitro of poliomyelitis viruses. IV. Viral multiplication in a stable strain of human malignant epithelial cells (strain HeLa) derived from an epidermoid carcinoma of the cervix

Effects of paramyxoviral infection on airway epithelial cell Foxj1 expression, ciliogenesis, and mucociliary function

Culture of Animal Cells: A Manual of Basic Technique

Preclinical models of lung cancer: cultured cells and organ culture

Human papillomavirus and the development of cervical cancer: concept of carcinogenesis

Relationship of alveolar epithelial injury and repair to the induction of pulmonary fi brosis

Generation of human pulmonary microvascular endothelial cell lines

Promotion of mitochondrial membrane complex assembly by a proteolytically inactive yeast Lon

Culture and transformation of human airway epithelial cells

A mechanism of hyperthermia-induced apoptosis in ras-transformed lung cells

Smoke/burn injury-induced respiratory failure elicits apoptosis in ovine lungs and cultured lung cells, ameliorated with arteriovenous CO 2 removal

Clonal growth of epithelial cells from normal adult human bronchus

Experimental systems for mechanistic studies of toxicant induced lung infl ammation

Regulation of vasodilator synthesis during lung development

Markers of airway smooth muscle cell phenotype

Airway smooth muscle cell proliferation: characterization of subpopulations by sensitivity to heparin inhibition

Divergent differentiation paths in airway smooth muscle culture: induction of functionally contractile myocytes

Airway wall remodelling and hyperresponsiveness: modelling remodelling in vitro and in vivo

NCI series of cell lines: an historical perspective

Establishment and characterisation of cell lines from patients with lung cancer (predominantly small cell carcinoma)

Cell culture methods for the establishment of the NCI series of lung cancer cell lines

Extracellular matrix proteins protect small cell lung cancer cells against apoptosis: a mechanism for small cell lung cancer growth and drug resistance in vivo

Radioimmunoassay of pulmonary surface-active material in the tracheal fl uid of the fetal lamb

Selected anatomic burn pathology review for clinicians and pathologists

Overview of ventilatorinduced lung injury mechanisms

Surfactant protein-A levels in patients with acute respiratory distress syndrome

Positive end-expiratory pressure modulates local and systemic infl ammatory responses in a sepsis-induced lung injury model

Injurious mechanical ventilation and end-organ epithelial cell apoptosis and organ dysfunction in an experimental model of acute respiratory distress syndrome

Alveolar infl ation during generation of a quasi-static pressure/volume curve in the acutely injured lung

Evaluation of mechanical ventilation for patients with acute respiratory distress syndrome as a result of interstitial pneumonia after renal transplantation

Primary cell culture of human type II pneumonocytes: maintenance of a differentiated phenotype and transfection with recombinant adenoviruses

Collagen-coated cellulose sponge: three dimensional matrix for tissue culture of Walker tumor 256

Histophysiologic gradient culture of stratifi ed epithelium

The three-dimensional question: can clinically relevant tumor drug resistance be measured in vitro?

Structural biology of epithelial tissue in histophysiologic gradient culture

Tissue culture in relation to growth and differentiation

The growth, development and phosphatase activity of embryonic avian femora and end-buds cultivated in vitro

Organ culture of mucosal biopsies of human small intestine

Growth of nodular carcinomas in rodents compared with multicell spheroids in tissue culture

Dialysis fermentor systems for concentrated culture of microorganisms

Semipermeable microcapsules

Application of ultrafi ltration for enzyme retention during continuous enzymatic reaction

Cell culture on artifi cial capillaries: an approach to tissue growth in vitro

Membranes and bioreactors: a technical challenge in biotechnology

Membrane bioreactors: engineering aspects

Membrane bioreactors: present and prospects

Artifi cial membrane as carrier for the immobilization of biocatalysts

Hollow fi ber cell culture in industry

Hollow fi ber enzyme reactors

Engineering of osteochondral tissue with bone marrow mesenchymal progenitor cells in a derivatized hyaluronan-gelatin composite sponge

Haemopoietic long-term bone marrow cultures from adult mice show osteogenic capacity in vitro on 3-dimensional collagen sponges

Microcarrier/cultures of animal cells

Mammalian cell culture: engineering principles and scale-up

Morphologic differentiation of colon carcinoma cell lines HT-29 and HT-29KM in rotating-wall vessels

Hydrodynamic effects on animal cells in microcarrier cultures

Physical mechanisms of cell damage in microcarrier cell culture bioreactors

Growth and death in over agitated microcarrier cell cultures

Cell death in the thin fi lms of bursting bubbles

Growth of anchoragedependent cells on macroporous microcarriers

Viscous reduction of turbulent damage in animal cell culture

Biological experiences in bubble-free aeration system

Critical parameters in the microcarrier culture of animal cells

The large-scale cultivation of mammalian cells

Analysis of gravity-induced particle motion and fl uid perfusion fl ow in the NASA-designed rotating zero-head-space tissue culture vessel

Cell culture for threedimensional modeling in rotating-wall vessels: an application of simulated microgravity

Autocrine and paracrine roles for growth factors in melanoma

Growth-regulatory factors for normal, premalignant, and malignant human cells in vitro

Reduced shear stress: a major component in the ability of mammalian tissues to form three-dimensional assemblies in simulated microgravity

Threedimensional culture of a mixed mullerian tumor of the ovary: expression of in vivo characteristics

Threedimensional modeling of T-24 human bladder carcinoma cell line: a new simulated microgravity vessel

Multicellular resistance: a new paradigm to explain aspects of acquired drug resistance of solid tumors

Comparison of cytotoxicity of agents on monolayer and spheroid systems

Modifi ed 2-tumour (L1210, colon 38) assay to screen for solid tumor selective agents

Signifi cance of three-dimensional growth patterns of mammary tissues in collagen gels

Assessing tumor drug sensitivity by a new in vitro assay which preserves tumor heterogeneity and subpopulation interactions

Factors affecting growth and drug sensitivity of mouse mammary tumor lines in collagen gel cultures

Origin and deposition of basement membrane heparan sulfate proteoglycan in the developing intestine

Tumor resistance to alkylating agents conferred by mechanisms operative only in vivo

Acquired multicellular-mediated resistance to alkylating agents in cancer

Heat sterilisation to inactivate AIDS virus in lyophilised factor VIII

Threedimensional growth patterns of various human tumor cell lines in simulated microgravity of a NASA bioreactor

Long term organ culture of human prostate tissue in a NASA-designed rotating wall bioreactor

Establishment of a three-dimensional human prostate organoid coculture under microgravity-simulated conditions: evaluation of androgen-induced growth and PSA expression

Three-dimensional coculture models to study prostate cancer growth, progression, and metastasis to bone

Threedimensional tissue assemblies: novel models for the study of Salmonella enterica serovar typhimurium pathogenesis

Series introduction: Innate host defense of the respiratory epithelium

The airway epithelium: structural and functional properties in health and disease

Apicobasal polarization: epithelial form and function

Robbins Infectious Diseases, 6th ed. Philadelphia: WB Saunders

Epithelial regulation of innate immunity to respiratory syncytial virus

Epithelia cells as regulators of airway infl ammation

Human bronchial epithelial cells with integrated SV40 virus T antigen genes retain the ability to undergo squamous differentiation

Identifi cation and culture of human bronchial epithelial cells

Developing differentiated epithelial cell cultures: airway epithelial cells

Evaluation of anchorageindependent proliferation in tumorigenic cells using the redox dye alamar Blue

Airway epithelium and mucus: intracellular signaling pathways for gene expression and secretion

Disorganization of stroma alters epithelial differentiation of the glandular stomach in adult mice

Expression and function of cell surface extracellular matrix receptors in mouse blastocyst attachment and outgrowth

Milk protein expression and ductal morphogenesis in the mammary gland in vitro: hormone-dependent and -independent phases of adipocyte-mammary epithelial cell interaction

Cell replication of mesenchymal elements in adult tissues. I. The replication and migration of mesenchymal cells in the adult rabbit dermis

Defi ning conditions to promote the attachment of adult human colonic epithelial cells

Laminin expression in colorectal carcinomas varying in degree of differentiation

Infl uence of mammary cell differentiation on the expression of proteins encoded by endogenous BALB/c mouse mammary tumor virus genes

Opinion: Building epithelial architecture: insights from three-dimensional culture models

Threedimensional engineered high fi delity normal human lung tissue-like assemblies (TLA) as targets for human respiratory virus infection

Severe acute respiratory syndrome (SARS)-CoV infection in a threedimensional human bronchial-tracheal (HBTE) tissue-like assembly