SEARCH

SEARCH BY CITATION

Keywords:

  • Betta splendens;
  • Bacillus cereus;
  • Polynuclear aromatic hydrocarbons

Abstract

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. COMPONENTS OF THE MICROCYTOSENSOR DEVICE
  5. CYTOSENSOR DATA ANALYSES
  6. SELECTED EXAMPLES OF CHROMATOPHORE RESPONSE TO TOXIC AGENTS
  7. FUTURE DIRECTIONS
  8. References

Fish chromatophores from Betta splendens are used as the cytosensor element in the development of a portable microscale device capable of detecting certain environmental toxins and bacterial pathogens by monitoring changes in pigment granule distribution. The adaptation of chromatophores to a microscale environment has required the development of enabling technologies to produce miniaturized culture chambers, to integrate microfluidics for sample delivery, to miniaturize image capture, and to design new statistical methods for image analyses. Betta splendens chromatophores were selected as the cytosensor element because of their moderate size, their toleration of close contact, and most importantly, for their responses to a broad range of chemicals and pathogenic bacteria. A miniaturized culture chamber has been designed that supports chromatophore viability for as long as 3 months, and that can be easily transported without damage to the cells. New statistical methods for image analyses have been developed that increase sensitivity and also decrease the time required for detection of significant changes in pigment granule distribution. Betta chromatophores have been tested for their responses to selected pathogenic bacteria and chemical agents. We discuss in detail the aggregation of pigment granules seen when chromatophores are incubated with Bacillus cereus, a common cause of food poisoning. Also described are the more subtle responses of chromatophores to a class of environmental chemical toxins, polynuclear aromatic hydrocarbons. We show that the chromatophores are able to detect the presence of certain polynuclear aromatic hydrocarbons at concentrations lower than the Environment Protection Agency (EPA) 550.1 standards.


Abbreviations
EPA

Environmental Protection Agency

PAH

polynuclear aromatic hydrocarbons

INTRODUCTION

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. COMPONENTS OF THE MICROCYTOSENSOR DEVICE
  5. CYTOSENSOR DATA ANALYSES
  6. SELECTED EXAMPLES OF CHROMATOPHORE RESPONSE TO TOXIC AGENTS
  7. FUTURE DIRECTIONS
  8. References

The ability to detect potential environmental toxins such as heavy metals, pesticides, organic chemicals and pathogenic bacteria is important in monitoring possible contamination. The development of portable devices capable of testing for environmental toxins in the field would be very useful in quickly identifying contaminated ground, water, or food. A multidisciplinary group involving the Departments of Biochemistry and Biophysics, Microbiology, Bioengineering, Industrial and Manufacturing Engineering, Electrical and Computer Engineering, Mechanical Engineering and Chemical Engineering are collaborating in the development of a portable microscale cytosensor device capable of analysing the environment for toxic chemicals and microbial pathogens. This device utilizes the chromatophores of the Siamese Fighting Fish, Betta splendens, as the cytosensor element. Adapting the chromatophores to a portable microscale device has required the development of enabling technologies to produce miniaturized chambers containing chromatophores, to integrate microfluidics for sample delivery, and to design statistics programs capable of quickly detecting and analysing various changes in chromatophore pigment granule distribution (1).

Chromatophores are the neurone-like cells containing pigment granules that are responsible for the brilliant colours of fish, amphibians, reptiles and cephalopods. Many animal species are capable of changing their colour as an adaptive behavioural response under the control of the nervous and endocrine systems, mediated by receptors on the surface of the chromatophores. Summaries of the many types of chromatophores and the variations in responses observed in different organisms are available in a number of reviews (2, 3). The specific types of chromatophores present in B. splendens and the changes in pigment granule distribution observed upon exposure to various agents will be described later. At the cellular level, colour changes in chromatophores can be produced in two different ways. Melanophores (brown to black), erythrophores (orange to red), and xanthophores (yellow to orange) can regulate their colour by having the pigment granules dispersed throughout the cytoplasm (dark) or by aggregating the pigment granules into a tight cluster in the centre of the cell (pale). Iridophores change their colours by an entirely different mechanism. They contain organelles consisting of stacks of purine crystals at very precise spacing designed to reflect light of a particular wavelength. Some types of iridophores can alter this spacing in response to neurotransmitters or other stimuli and thus change the wavelengths of light reflected (2, 3).

Chromatophores contain various types of neurotransmitter and hormone receptors, depending on the species. The most important signal transduction mechanism linking agent receptor binding with pigment aggregation or dispersion is the classic G-protein-linked type. Receptors that cause aggregation are linked to the Gi, whose activation results in a decrease in cAMP, while receptors that cause dispersion are linked to Gs, whose activation results in an increase in cAMP. This cAMP increase activates cAMP-dependent protein kinase (protein kinase A), which phosphorylates and activates other proteins, initiating a cascade of events resulting in pigment granule dispersion. The long-range movements of pigment granules depend on polar microtubules and specific motor proteins bound to the pigment granules. There are two major families of motor proteins, kinesins that move their cargo outward, and dyneins that move the granules inward towards the centrosome. The ability of these motor proteins to bind microtubules and thus transport pigment granules is regulated by the phosphorylations resulting from the signal cascade initiated by G-protein-linked receptor binding (3–5). Although G-protein-linked receptors have received the most attention, there is evidence that Ca++ can also affect pigment granule movement in chromatophores containing β2 adrenergic receptors. In addition, shorter-range granule movements have recently been shown to depend on the interaction of myosin V and actin, and that this interaction is controlled by phosphorylation of myosin V mediated by calcium/calmodulin-dependent kinase II (6–8).

Chromatophores have been an extremely useful reporter system to study G-protein-linked receptor/ligand interactions, as a tool for screening pharmacological agents and for dissecting the structure/function relationships of various ligands. However, in developing a broad-based cytosensor, we want to make use of the complexity of the signal transduction cascades affecting both long-range and short-range granule movements. This complexity provides a powerful tool to detect the effects of agents that might perturb the pathway at a number of different points, as well as agents that interfere more directly with microtubule transport and general cell metabolism. Inherent in the system is the ability to monitor not only direct effects to perturb the normal state of the chromatophores, but also the ability to challenge the chromatophores with various agents and observe differences in the expected response. Signal amplification can also be a source of increased sensitivity to certain agents. For example, a particular agent might completely inhibit aggregation, slow it down, or cause some sort of partial aggregation, depending on where it acted. Agents that interfere with glycolysis or the metabolism of ATP could affect granule movement by reducing the pool of ATP required for motor protein movement. Cross talk among the signal-transduction pathways can also increase the repertoire of agents to which the cells are sensitive. Also, there is increasing evidence that macromolecular signalling complexes are formed within cells that can give rise to chemical compartmentalization, an important consideration when observing large dendritic chromatophores (9). Our observations of the varied responses of Betta chromatophores to certain toxins and pathogenic bacteria have led to the design of novel computer algorithms for image analysis, and we have begun to catalogue these responses. Our plan is to be able to develop categories of responses common to groups of agents that target specific components of the signal transduction cascade. This would permit an assignment of a cytosensor response to a particular class of agents. In the following sections we describe the current state of development of the microscale cytosensor device. We are continually creating new and improved prototypes, and testing other environmental toxins and pathogenic bacteria. As we progress, we have become increasingly more encouraged about the potential usefulness of this system.

COMPONENTS OF THE MICROCYTOSENSOR DEVICE

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. COMPONENTS OF THE MICROCYTOSENSOR DEVICE
  5. CYTOSENSOR DATA ANALYSES
  6. SELECTED EXAMPLES OF CHROMATOPHORE RESPONSE TO TOXIC AGENTS
  7. FUTURE DIRECTIONS
  8. References

Choice of Betta splendens and Development of Primary Tissue Culture Protocol

Many species of fish contain motile chromatophores, including the popular medaka, angelfish, goldfish, zebrafish and various African cichlids such as the tilapia and jewel cichlid. To our knowledge, we are the first group to systematically investigate the chromatophores of B. splendens. Fish chromatophores are terminally differentiated and do not replicate in tissue culture. Although this might seem at first to be a disadvantage, we have found that by inhibiting the growth of other replicating cell types present in the primary tissue culture, the chromatophores remain viable up to 3 months (10). Their energy requirements in the resting state appear to be small, so that transporting them in sealed containers with minimal media change is possible. In contrast, cells that are dividing require subculture at regular intervals and require more frequent media changes. This puts serious constraints on their use in a field device where a long ‘shelf life’ is important. The main disadvantage in using primary cell cultures to populate the cytosensor is that different fish are used each time, and there are fish-to-fish variations in the various subpopulations of chromatophore types. There may be also individual variations in the responses to various agents. We are in the process of extensive calibration experiments to document the fish-to-fish variation and to develop internal controls that will permit us to standardize the responses.

In order to use chromatophores as cytosensors, the development of a primary tissue culture protocol was essential. We have developed such a protocol, isolated the chromatophores from a large number of these species, and compared their culture characteristics (1). While it is beyond the scope of this review to provide detailed comparisons, some general comments can be made about our choice of B. splendens as the cytosensor. This choice was driven by several requirements for the device. In order to obtain enough cells for a statistically valid sample within a small area in a miniaturized device, relatively small chromatophores that tolerate close contact were needed. In addition, the device required a chromatophore type that could survive for at least a month. These characteristics were found in the erythrophores and melanophores of B. splendens, but not in other species. For example, 1000 B. splendens chromatophores can be cultured in an area of less than 1 mm2. In contrast, the same area would only contain about 10 tilapia chromatophores, not only because of their larger size, but also because they fail to thrive at high densities. These and other failings were found in the other tested species of fish (1). Betta splendens has many advantages that we have exploited. The fish come in a variety of colours easily obtainable from commercial sources. This allows us to select fish enriched in various types of chromatophores, although at this point we are concentrating on the erythrophores for device development. The fins are densely packed with chromatophores, giving them their vivid colours. In tissue culture the chromatophores tend to recreate their close proximity in the fin, providing the high cell densities important for optimal device function. Displayed in Fig. 1 are some examples of the variety of colours of Bettas, as well as two examples of primary tissue cultures, one of which is enriched in erythrophores and the other in melanophores.

image

Figure 1.  The right hand panel shows the characteristics of primary cell cultures enriched in erythrophores (1) and melanophores (2). Note the varied morphology and colour intensity within each of these cultures. The left panel illustrates some examples of selected colours of B. splendens enriched in erythrophores (A), and melanophores (B). (Pictures courtesy of A. C. Shaffer, Pumkinnpie’s Betta Sales). The density and vivid colour intensities of chromatophores in the fins are illustrated in (C).

Download figure to PowerPoint

We have also started to explore the usefulness of iridophores as biosensors. Their development has lagged behind other types of chromatophores for several reasons. First, if a Betta is selected that appears to have uniform iridescence, in primary tissue culture the individual cells are a rainbow of colours. These colours are combined in the living fish to produce the uniform colour of iridescence that is seen. If these cells are exposed to an agent such as noradrenaline, only a small fraction (10–20%) will react by increasing the wavelength of reflected light. We are currently testing many different colours of Bettas in an attempt to find some that have a higher proportion of reactive iridophores. The second approach is to investigate the iridophores from other species of fish. We have found that the iridophores of the marble angelfish, Pterophyllum scalare, have provided the best system to date for developing iridophores as biosensors. As seen in Fig. 2, these iridophores are densely packed and contain a high percentage of reactive cells.

image

Figure 2.  Cultured iridophores from the angelfish (Pterophyllum scalare). The iridophores were illuminated at an angle using a fibre-optic light source. The reflected colours of the iridophores are shown before (A) and 15 min after the application of 1 μM noradrenaline (B).

Download figure to PowerPoint

Miniaturization

Crucial to the development of a portable field device is the miniaturization of the components. In practical terms, we need to integrate and co-ordinate the efforts of the eight different groups participating in various components of this effort. We are currently in the fourth iteration of device development and the design is shown in Fig. 3. The chamber design allows population with the chromatophores through one of the ports, and the cells are cultured in an area approximately 3 mm long by 2 mm wide and approximately 250 μm deep. The ports on either side are approximately 2 mm from the interior part of the chamber containing the cultured cells. Testing of various configurations has shown that the chromatophores easily tolerate these confined conditions with relatively little opportunity for diffusion of gases and nutrients. We performed a test in which chromatophores were cultured in these chambers and the media inside the chamber was replaced just once per week. They remained healthy and responsive for 3 months (10). The Betta chromatophores are extremely sticky, and grow best on untreated polycarbonate. We found that coating the surfaces with various types of cellular adhesion proteins, such as collagen IV and fibronectin, actually reduced the health and longevity of the culture. Thus, the Betta chromatophores are very robust and adapt well to the constraints of miniaturized culture conditions. This version of the device uses wicks for sample delivery. This portable device includes an LED light source, lens, camera, monitor, keyboard, and a computer containing our specialized image processing software. The next iteration of the device will incorporate a pumping system for sample delivery, further miniaturization of the optics, and updated image processing capabilities.

image

Figure 3.  Cell-based biosensor prototype. The chamber containing chromatophores has wicks inserted for sample delivery, and is placed into the holder. An agent can be applied to the top wick, and the response of the chromatophores is imaged using the LED light source, lens and camera. The image is then processed by the statistical program present in the computer.

Download figure to PowerPoint

CYTOSENSOR DATA ANALYSES

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. COMPONENTS OF THE MICROCYTOSENSOR DEVICE
  5. CYTOSENSOR DATA ANALYSES
  6. SELECTED EXAMPLES OF CHROMATOPHORE RESPONSE TO TOXIC AGENTS
  7. FUTURE DIRECTIONS
  8. References

Our experience strongly suggests that important factors in the successful determination and assessment of chemical and biological threats with the cytosensor system are the appropriate choice of observation granularity and problem specific averaging through the use of appropriate models (11). Modern measurement techniques enable observations of clusters of cells, individual cells, or subcellular features. These levels of observation granularity are directly related to various embedded levels of averaging. For example, cell clusters carry information already averaged over cells of interest, as well as over the surrounding elements, such as growth media or cells of no experimental interest, whereas individual cells carry information about subcellular activities precipitated by exposure to extracellular stimuli. For a given organism, the initial biological reaction occurs at a certain level of granularity, and with the passage of time ‘spreads’ and causes changes at higher granularity levels. Eventually, the change is observed at the level of the entire organism, being seen as impaired function or death. If the granularity level is set too high for initial observations, then biological reactions may appear to be nonspecific or may not appear at all within the timeframe of interest. A properly chosen granularity for observation is therefore crucial for the early or timely detection of threats (12).

The need for statistical methods to analyse the data generated by cytosensor systems is quite obvious. Model based data analysis is the preferred method when dealing with complex and ambiguous data sets, such as those produced with the cytosensor system. Furthermore, missing data and dropouts are factors that cannot be avoided or ignored because of the nature of the experiments that are performed with this system. Highly complex data such as this dictates that model validation be done differently from conventional statistics, where models are generic in the sense that they may be applied to a wide class of problems. Applied statistics books tend to use a small number of example data sets as illustrations of all presented methods, and many statisticians often find outliers, missing data and dropouts to be ‘nuisance’ factors, complicating the analysis. This reflects the statisticians’ belief in the generality of their methods and limits the versatility and effectiveness of these methods to adapt to the problem at hand. We recognize that the assumptions that underlie conventional statistics (various independence conditions, sample uniformity or distribution) are rarely met with the data from the cytosensor system and that outliers should be integrated into any working model of system response. Indeed, generic statistical methods like analysis of variance (ANOVA) are capable of eliciting only a small part of the structured information contained within the cytosensor data because outliers provide much of the information (13).

Finally, given the ambiguous nature of the data sets generated by the cytosensor system, system outputs should reflect a distribution of output decisions rather than crisp numerical outputs. These distributions should take the form of relative weights expressing the sensor ‘confidence’ in different experiment outcome scenarios. Such algorithms have been devised using the concept of probabilistic experts, themselves a generalization of probabilistic risk analysis methods (11).

These statistical approaches are being developed for two main purposes. The first is to increase the sensitivity of detecting a change in pigment granule distribution when the chromatophores are exposed to an agent. The cytosensor contains a heterogeneous population of cells, and it is frequently observed that a small percentage of chromatophores respond much earlier than others, some respond more slowly, and some do not respond at all. The ability to determine if a significant dynamic change has occurred in this smaller population of fast-responding cells can potentially provide improved sensitivity. The second purpose is to detect more subtle changes in pigment granule distribution, and to determine if particular classes of toxic agents have similar effects or ‘signatures’ that can be exploited. We are in the process of assembling a large database of responses to various agents for future development of the statistical methods. The selected examples discussed below illustrate the problems we face in statistical analyses, and will provide context for this theoretical overview.

SELECTED EXAMPLES OF CHROMATOPHORE RESPONSE TO TOXIC AGENTS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. COMPONENTS OF THE MICROCYTOSENSOR DEVICE
  5. CYTOSENSOR DATA ANALYSES
  6. SELECTED EXAMPLES OF CHROMATOPHORE RESPONSE TO TOXIC AGENTS
  7. FUTURE DIRECTIONS
  8. References

General Comments

Chromatophores have been used extensively as a tool to study specific responses to a wide range of pharmacologic and bioactive agents and their analogues in a number of different species. We are using chromatophores as biosensors to detect a broad range of bioactive compounds, without necessarily knowing the exact mechanisms inducing changes in the cells. Table 1 provides an overview of several classes of agents that have been shown to affect chromatophore pigment distribution. Although our ultimate goal is to understand the cell biology behind these changes, our primary goal now is to catalogue the various types of changes we observe, develop a library of these responses, and see if any correlations develop among different categories of agents. We have seen such a pattern in the response of chromatophores to certain organophosphate pesticides. When exposed to these agents, the chromatophores initially show varying degrees of partial aggregation and uneven distribution of pigment granules. However, most striking is their impaired ability to aggregate pigment granules when challenged with noradrenaline (1). Organophosphate pesticides act by inhibiting acetylcholinesterase resulting in the build-up of acetylcholine at synapses. However, the Betta chromatophores do not produce acetylcholine, and the toxic effect is believed to be caused by the inhibition of an uncharacterized esterase (1). However, no such clear pattern has emerged in our experiments testing certain heavy metals. They have a variety of effects on chromatophore pigment granule distribution, with some causing hyperdispersion, some causing partial aggregation, and others having no visible effects (1).

Table 1.   Classes of environmental toxins that cause changes in pigment granule distribution Thumbnail image of

Bacillus cereus

The detection of pathogenic bacteria in the environment is a major concern. There are serious public health issues concerning bacterial contamination of food and water. A major focus of our group is investigating the use of chromatophores as cytosensors for pathogenic bacteria. Table 1 lists some pathogenic bacteria and bacterial toxins that have caused measurable changes in Betta chromatophores. We have tested nonpathogenic bacteria, such as Lactococcus lactis, Bacillus subtilis, and Escherichia coli JM101, and these bacteria did not cause any noticeable change under the same experimental conditions as the pathogenic bacteria (1). Toxins produced by bacteria have been widely studied, and there are a number of sensitive specific tests for the presence of particular bacteria or their toxins. Some of the best known are the ADP-ribosylating toxins seen in diphtheria and cholera; neurotoxins seen in tetanus and botulism, and membrane-damaging toxins, such as produced by Staphylococcus aureus. However, most bacteria produce more than one type of toxin, as discussed below using B. cereus as an example, and a cytosensor capable of detecting a broad range of bacterial toxins has potential as an initial screening tool.

Bacillus cereus, a Gram-positive spore-forming bacterium, has been shown to produce four or more enterotoxins that are responsible for two different types of food poisoning and other infections (14). The diarrheal types of food poisoning have been attributed to the production of a group of heat-labile protein complexes (haemolysin BL and nonhaemolytic enterotoxin) that have been postulated to form pores in the membrane of target cells (15). Abdominal pain and diarrhoea ensue after 8–16 hr incubation.

The emetic or vomiting syndrome appears within 1–5 hr after ingestion and is caused by a dodecadepsipeptide called cereulide; a unique enterotoxin that is resistant to proteolytic degradation, pH extremes and high temperature (survives 121°C for 90 min) (16). Symptoms elicited by the toxin in a bioassay based on loss of motility in boar spermatozoa were similar to those caused by the potassium ionophores valinomycin and gramicidin (swollen mitochondria and a blockage of oxidative phosphorylation) (17).

In addition, a new cytotoxin isolated from a B. cereus strain associated with a severe food poisoning outbreak in France was found to have sequence similarity to a family of β-barrel channel-forming toxins including Clostridium perfringens β-toxin and S. aureus haemolysins. The strain responsible for this outbreak caused a bloody diarrheal syndrome from which three people died. Interestingly this strain did not contain the genes for any of the other known B. cereus enterotoxins (18).

Betta erythrophores incubated with a culture of B. cereus bacteria or with the filtered supernatant responded by aggregating their pigment granules as shown in Fig. 4. We are investigating which toxin or combination of toxins is responsible for the aggregation of the pigment granules, and attempting to determine the mechanisms responsible for pigment granule aggregation.

image

Figure 4.  Response of erythrophores to exposure to Bacillus cereus. Cultures were grown aerobically in media known to enhance production of specific toxins. Bacterial cells plus culture supernatant were applied to the chromatophores and allowed to react for an hour. The appearance of the chromatophores before (A) and after exposure (B) are shown.

Download figure to PowerPoint

Polynuclear Aromatic Hydrocarbons

Polynuclear aromatic hydrocarbons (PAHs) are produced when materials containing carbon and hydrogen are burned. The primary sources of production are by internal combustion engines and the burning of fossil fuels. Inhalation of PAHs through cigarette smoke is probably the most serious source of exposure, as some PAHs, such as benzo(a)pyrene, are classed as known animal carcinogens and probable human carcinogens, and most PAHs containing three or more rings are genotoxic (19). They are resistant to biological degradation and tend to persist in contaminated environments (20). Current US EPA methods for PAH analysis (e.g. Method 550.1) require approximately 2 d.

We have tested the ability of the chromatophores to respond to various PAHs. These compounds have very low water solubility, and thus the bioavailability of these compounds may differ from their concentration in water or ground sediments. Tested PAHs included benzo(a) pyrene, 1.2:5.6-dibenzeneanthracene, phenanthrene, fluorine, 1.12-benzoperylene, anthracene, acenaphthene, benzo(k) fluoranthene, benzo(b)fluoranthene, chrysene, naphthalene, indeno (1.2.3-C.D.) pyrene. These compounds were diluted in Tris-buffered saline and the chromatophores exposed to these agents for 3 h. Images of chromatophores before and after exposure were processed using ImageProTM software and the areas occupied by pigment granules calculated. As shown in Fig. 5, the change measured was a partial or complete aggregation of the pigment granules in some of the cells. We used an arbitrary 10% decrease in cell area to define a response to a particular PAH. The images shown in Fig. 5 were recorded using polarized light in order to see if the response was the result of a true aggregation, or whether the erythrophores had shrunk. The presence of the cellular outlines show that partial aggregation had occurred. Robust changes in cytosensor response to chrysene, naphthalene, and indeno (1.2.3-C.D.) pyrene were observed at 1 μg/l. Therefore, the cytosensor bioassay achieved a 6-order of magnitude increase in measurement sensitivity relative to older methodologies. Comparison of the screened compounds shows that seven PAHs elicited a response at the initial screening concentration (1000 μg/l) while nine gave indeterminate responses. Of the seven responders, four were examined in detail – pyrene, 1,2-benzanthracene, fluoranthene and acenaphthene (Table 2). Pyrene and acenaphthene were detectable below the mandated detection limits per EPA 550.1. Fluoranthene, although not as strong a responder, was still detectable at its mandated detection limit, while 1,2-benzanthracene was detectable well above its EPA mandated detection limit. Responses to the remaining PAHs in this listing were qualitatively difficult to observe.

image

Figure 5.  Response of erythrophores to exposure to acenaphthene. The chromatophore culture was exposed to 0.1 mg/ml acenaphthene for a period of 3 h. The appearance of the cells before (A) and after treatment (B) is shown. These images taken using polarized light show that shrinkage of the cells has not occurred, and that there is a change in pigment granule distribution. A number of the chromatophores exhibit marked aggregation (arrows), while other cells show a more modest response, or none at all.

Download figure to PowerPoint

Table 2.   Comparison of EPA 550.1 and cytosensor sensitivity Thumbnail image of

It is important to note that not all of the cells exhibited a strong response, as readily seen in Fig. 5. A heterogeneous population of chromatophores is always present in primary tissue culture, and not all cells respond equally when exposed to many of the environmental toxins tested. The new statistical approaches to image analyses discussed earlier are designed to provide more sensitive results in cases where only a fraction of the cells show a strong response. We are in the process of using the PAHs to refine the new statistical approach to further enhance the sensitivity of detection.

These results are exciting because they demonstrate a greater sensitivity to a class of agents than that seen with current established methodology. The fact that pyrene, a PAH standard, elicits a strong response is noteworthy. Further, unlike the EPA mandated detection limits per EPA 550.1, the cell response is indicative of biological impairment and thus directly linked to chemical bioavailability. Thus, these results suggest that acenaphthene is toxic at levels far below the mandated detection limit, whereas 1,2-benzanthracene is only toxic at levels far above the mandated EPA detection limit. Future research efforts will involve paired studies designed to determine whether impairments in cell response correlate with the bioavailability of PAHs in the environment.

FUTURE DIRECTIONS

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. COMPONENTS OF THE MICROCYTOSENSOR DEVICE
  5. CYTOSENSOR DATA ANALYSES
  6. SELECTED EXAMPLES OF CHROMATOPHORE RESPONSE TO TOXIC AGENTS
  7. FUTURE DIRECTIONS
  8. References

In this review, we describe using fish chromatophores as a broad-based cytosensor to detect a broad range of environmental toxins and pathogenic bacteria. Our goal is to continue development of our portable field device by increased miniaturization of the components. Another major goal is the testing of other classes of toxins in the hope of broadening the range of agents that chromatophores can detect. Continued development of the new statistical approaches will result in more sensitive detection of toxic agents, and hopefully permit the extraction of information from more subtle changes in pigment granule distribution that are beyond our capabilities now. The development of a portable device capable of taking an environmental sample, detecting its effects on chromatophores, and then interpreting the results in a timely fashion requires an enormous effort utilizing the skills of both biologists and engineers. The basic cell biology underlying the varied responses of the chromatophores to various agents needs to be advanced in order to properly interpret the effects of real-life samples containing an unknown agent. Advances in combining micromachining, microfluidics, optics and novel statistical software will be required to achieve our goal.

Footnotes
  1. *Address reprint requests to Rosalyn H. Upson, Department of Bioengineering, 116 Gilmore Hall, Oregon State University, Corvallis, OR 97331, USA. E-mail: upsonr@engr.orst.edu

References

  1. Top of page
  2. Abstract
  3. INTRODUCTION
  4. COMPONENTS OF THE MICROCYTOSENSOR DEVICE
  5. CYTOSENSOR DATA ANALYSES
  6. SELECTED EXAMPLES OF CHROMATOPHORE RESPONSE TO TOXIC AGENTS
  7. FUTURE DIRECTIONS
  8. References
  • 1
    McFadden P, Chaplen F, Kolodziej W, Jovanovich G, Plant T, Liburdy J, Paul B, Trempy J, Gerwick W, Willard C. Methods of Detecting Bioactive Compounds. US Patent and Trademark Office, Patent Application; 2001
  • 2
    Fujii R. Cytophysiology of fish chromatophores. Internat Rev Cytol 1993;143:191255
  • 3
    Fujii R. The regulation of motile activity in fish chromatophores. Pigment Cell Res 2000;13:300319
  • 4
    Reese E & Haimo L. Dynein, dynactin, and kinesin II’s interaction with microtubules is regulated during bidirectional organelle transport. J Cell Biol 2000;151:155166
  • 5
    Nilsson H. Melanosome and erythrosome positioning regulates cAMP-induced movement in chromatophores from spotted triplefin, Grahamina capito. J Exp Zool 2000;287:191198
  • 6
    Brown SS. Cooperation between microtubule- and actin-based motor proteins. Annu Rev Cell Dev Biol 1999;15:6380
  • 7
    Westbroek W, Lambert J, Naeyaert JM. The dilute locus and Griscelli syndrome: gateways towards a better understanding of melanosome transport. Pigment Cell Res 2001;14:320327DOI: 10.1034/j.1600-0749.2001.140503.x
  • 8
    Tuma MC & Gelfand VI. Molecular mechanisms of pigment transport in melanophores. Pigment Cell Res 1999;12:283294
  • 9
    Karpen JW & Rich TC. The fourth dimension in cellular signaling. Science 2001;293:22042205DOI: 10.1126/science.293.5538.2204
  • 10
    Aramphongphun C. Application of ultrasonic welding to the rapid prototyping of microfluidic systems for biotechnology. MS Thesis. Oregon: Oregon State University; 2001
  • 11
    Kolodziej W, Pacut A, Chaplen F. Statistical signal processing for cytosensors. In: 4th International Conference ‘Neural Networks and Expert Systems in Medicine and Healthcare’, Milos Island, Greece; 2001
  • 12
    Pacut A. Stochastic Modeling of Diverse Scales: From Poisson to Network Neurons. Warsaw, Poland: Oficyna Wydawnicza PW; 2000
  • 13
    Kolodziej W, Pacut A, Chaplen F, Upson R, Roussel N. Statistical Approach for Response Detection from Sensor Arrays. US Patent and Trademark Office, Provisional Patent Application; 2001
  • 14
    Jay JM. Modern Food Microbiology. New York: Van Norstrand Reinhold; 1994. pp. 501–503
  • 15
    Lund T & Granum PE. Comparison of biological effect of the two different enterotoxin complexes isolated from three different strains of Bacillus cereus. Microbiology 1997;143:33293336
  • 16
    Agata N, Ohta M, Mori M, Isobe M. A novel dodecadepsipeptide, cereulide, is an emetic toxin of Bacillus cereus. FEMS Microbiol Lett 1995;129:1720
  • 17
    Andersson MA, Mikkola R, Helin J, Andersson MC, Salkinoja-Salonen M. A novel sensitive bioassay for detection of Bacillus cereus emetic toxin and related depsipeptide ionophores. AEM 1998;64:13381343
  • 18
    Lund T, DeBuyser M, Granum PE. A new cytotoxin from Bacillus cereus that may cause necrotic enteritis. Mol Microbiol 2000;38: 254261
  • 19
    Rodgman A, Smith CJ, Perfetti TA. The composition of cigarette smoke: a retrospective with emphasis on polycyclic components. Hum Exp Toxicol 2000;19:573595
  • 20
    Arzayus KM, Dickhut RM, Canuel EA. Fate of atmospherically deposited aromatic hydrocarbons (PAHs) in Chesapeake Bay. Environ Sci Techol 2001;35:21782183