Final Report: Development and Evaluation of Procedures for Detection of Infectious Microsporidia in Source Waters

EPA Grant Number: R828042
Title: Development and Evaluation of Procedures for Detection of Infectious Microsporidia in Source Waters
Investigators: Rochelle, Paul A. , Johnson, Anne M. , Leitch, Gordon , Visvesvara, Govinda
Institution: Metropolitan Water District of Southern California , Centers for Disease Control and Prevention , Morehouse School of Medicine
EPA Project Officer: Nolt-Helms, Cynthia
Project Period: May 1, 2000 through May 1, 2002
Project Amount: $294,635
RFA: Drinking Water (1999) RFA Text |  Recipients Lists
Research Category: Drinking Water , Water


The overall objectives of this research project were to: (1) develop methods to recover microsporidia from water; (2) determine the viability and infectivity of detected spores; and (3) use the methods to assess the occurrence of microsporidia in untreated source waters. The specific objectives of this research project were to: (1) evaluate concentration and purification approaches for recovery of microsporidia spores from environmental water samples; (2) evaluate molecular and microscopic methods for detection of microsporidia; (3) optimize and compare in vitro viability and infectivity assays for microsporidia recovered from water; and (4) determine the prevalence of infectious microsporidia in natural water sources.

The microsporidia group of protozoa, particularly Enterocytozoon bieneusi and Encephalitozoon spp., are considered emerging pathogens that represent serious threats to public health. Some microsporidia can lead to disseminated infection affecting almost every organ of the human body in immune-compromised patients, and they also can infect immunocompetent individuals. Because many animals carry microsporidia, it is possible that surface waters can be contaminated and consequently may serve as a route of transmission to humans. Very little is known, however, about the occurrence of microsporidia in environmental water sources, and there is a critical need to determine the role that drinking water plays in the epidemiology of this group of parasites. Moreover, there are no routine methods for detection of microsporidia in water.

Summary/Accomplishments (Outputs/Outcomes):

A variety of procedures were evaluated for enumeration of microsporidia spores. Calcofluor is a fluorescent brightening agent that nonspecifically binds to chitin and cellulose. It is therefore useful for improving microscopic observations of fungi and other organisms, such as microsporidia, that contain chitin and/or cellulose in their cell walls. For enumeration of purified spore preparations and spores recovered from seeded, treated drinking water samples, spores were stained with 0.05 percent Calcofluor in dilute potassium hydroxide for 30 minutes and examined by epifluorescence microscopy with an excitation wavelength of 340-380 nm. Recognition of spores and speed of enumeration were improved by counterstaining with 0.2 percent Evans Blue for 30 minutes to reduce background fluorescence. Spores of Encephalitozoon spp. exhibited bright blue fluorescence. The size of E. intestinalis spores based on microscopic observation of Cellufluor-stained preparations was 2.6 ± 0.3 μm (mean ± standard deviation) × 1.2 ± 0.2 μm. Three fluorescein-labeled polyclonal antibodies directed against Encephalitozoon spp. also were evaluated. All antibodies demonstrated nonspecific staining with considerable background staining, particularly when spores were suspended in untreated source water samples. In addition, a substantial number of spores did not stain at all with these antibodies. When comparing hemacytometer enumeration of unstained E. intestinalis spores with spores enumerated by fluorescence microscopy, an average of 35 percent were detected by Calcofluor staining, and 43 percent were detected by staining with a fluorescein-labeled antibody. There was no significant difference in the proportion of the total spore population that stained when comparing Calcofluor to fluorescent antibody staining (α = 0.05, P = 0.37, N = 33). The average coefficient of variation when using fluorescent antibody staining to enumerate replicate spore doses applied to microscope slides was 22 percent.

Published PCR primers were evaluated for their sensitivity and specificity in detecting E. intestinalis, E. hellem, and E. cuniculi. Amplification specificity was assessed using DNA from target organisms and a wide variety of target organisms that may be present in untreated surface waters. Primers amplifying fragments of the small subunit ribosomal RNA (SSU rRNA) gene were specific for each of their respective target organisms. Primers specific for the β-tubulin gene amplified the intended target sequence from E. intestinalis and E. hellem . As an initial step in improving the reliability of PCR for detection of microsporidia, real-time assays were developed for E. intestinalis using a LightCycler instrument with direct fluorescence incorporation and fluorescent resonance energy transfer (FRET) technologies. Plotting the threshold cycle (Ct) against relative DNA concentration for a direct fluorescence assay demonstrated that the reaction was linear over the five orders of magnitude, and the amplification efficiency was 2.5 (perfect amplification efficiency = 2.0). A quantitative PCR assay also was developed using FRET technology and specific hybridization probes designed for this research project. FRET-based assays require two hybridization probes that are internal to the PCR primers and hybridize in a head-to-tail configuration. Each probe is labeled with a different marker dye and interaction of the two dyes can occur only when both are bound to their respective targets. Upon excitation by the appropriate wavelength of light, the donor fluorophore transfers energy to the adjacent acceptor fluorophore, which emits measurable light. The probes together with the SSU rRNA primers constitute a highly specific detection assay for E. intestinalis because hybridization of all four oligonucleotides to their respective targets is required for detection by the FRET assay. The assay was linear over five orders of magnitude and had an amplification efficiency of 1.9.

An improved method was developed to purify E. bieneusi spores from the stools of AIDS patients with microsporidiosis. Frozen stool samples obtained from E. bieneusi-infected patients were subjected to two rounds of density gradient centrifugation through 90 percent Percoll followed by 46 percent cesium chloride. Spores obtained from the second density gradient purification step were purified further using a cell sorter. The purity of the E. bieneusi spore preparations was assessed at each step of the procedure by staining with Calcofluor, immunofluorescent staining, and transmission electron microscopy. The cell sorting step significantly aided in the purification of spores and provided a preparation that could be used in the preparation of monoclonal and polyclonal antibodies against E. bieneusi. Specific antibodies are essential for immunofluorescent identification of microorganisms and for purification methods that rely on immunomagnetic separation.

A variety of filtration and concentration approaches were evaluated for recovery of microsporidia spores from spiked water samples. These included capsule filters in a range of porosities, flat membrane filters, compressed foam filters, centrifugation, and immunomagnetic separation. Commercially available capsule filters with an absolute porosity of 1 µm were evaluated for recovery of E. intestinalis spores from 10 L of untreated source water, along with custom-made capsule filters with porosities of 0.55 µm and 0.8 µm. The highest recovery efficiencies were obtained with the 0.55 µm capsule filter (37 ± 19%), but there was considerable variability in recoveries as evidenced by the relatively high coefficients of variation. The recovery efficiency with compressed foam filters with a nominal porosity of 1 μm was 6 percent for 291 E. intestinalis spores spiked into 11 L of treated drinking water compared to 19 percent for Cryptosporidium parvum oocysts. Smaller porosity filters also were manufactured by including more foam discs in each filter. The final concentrate from these filters, however, contained a lot of particulate matter (presumably from the filter materials) that interfered with recovery efficiencies. Initial trials with centrifugal filters consisting of a modified nylon membrane in a centrifuge tube involved recovering spiked E. intestinalis spores from 15 mL of treated drinking water (67 ± 8 spores). Recovery efficiencies of 52 percent, 14 percent, and 18 percent were obtained with centrifugal filters containing membranes with porosities of 0.2 µm, 0.3 µm, and 0.45 µm, respectively. Larger versions of these centrifugal devices are available to accommodate increased volumes of water for scaleup studies. Experiments with flat membrane filters demonstrated a threefold increase in spore retention by 0.1 mm porosity filters compared to 0.6 µm filters.

A prototype ultrafiltration device also was evaluated for recovery of microorganisms from large volumes of water. The apparatus contained a 68 kDa filter and was capable of concentrating 100 L of water down to a final volume of 100 mL (1,000-fold concentration) in less than 90 minutes. E. intestinalis spores were seeded into 50 L of treated drinking water (3.2 × 103 spores/mL), and recovered spores in the final concentrate were enumerated by epifluorescence microscopy of Calcofluor stained slides. The average recovery efficiency for E. intestinalis spores was 44 percent compared to 43 percent for similarly sized Bacillus subtilis spores.

Recovery efficiencies for an indirect immunomagnetic purification system ranged from 9.2-55 percent for 218-414 seeded spores with an average recovery efficiency of 37.5 percent. Based on a single factor analysis of variance of direct comparisons, there was no significant difference between immunomagnetic separation (IMS) recovery of E. intestinalis spores using paramagnetic beads conjugated to the two different antibodies (means of 33% and 31%, P = 0.86, N = 16). The majority of the IMS purification experiments were conducted using a 10 mL closed-tube format and external magnet for magnetic separation. A continuous flowing format also was evaluated in which the sample containing paramagnetic beads was passed repeatedly through a magnetic screen. Direct comparisons between the two formats demonstrated average recovery efficiencies of 37 percent for the external magnet format compared to 18 percent for the recirculating internal magnet format (spike = 414-474 spores).

A variety of cell lines (RK13, MDCK, BGMK, HCT-8, Caco-2, and MA104) were assessed for their ability to support in vitro infection with E. intestinalis. Infectivity was measured by inoculating E. intestinalis spores onto cell monolayers grown in 48-well culture plates and incubated at 35°C for 72 hours in a 5 percent CO2 atmosphere. Following incubation, infection was detected by extracting total cellular RNA (host cell and parasite) followed by RT-PCR with primers specific for E. intestinalis SSU rRNA. Inoculation of RK13 cells with 1,000 spores inactivated by heat (70 °C for 60 minutes) or 10 percent formalin (2 hours at room temperature) did not result in amplification by RT-PCR, thus demonstrating that the assay did not detect dead spores. Calcofluor staining also was evaluated for detecting infection in cell monolayers but was unreliable because of the high level of background fluorescence.

Primers targeting the SSU rRNA, hsp70, and β-tubulin genes were evaluated for RT-PCR detection of infection in cell cultures. Comparison of the results obtained with viable spores and spores that had been inactivated by heat (70 °C, 60 minutes) and incubation in 10 percent formalin for 2 hours indicated that the SSU rRNA primers yielded the most consistent results: positive detection with viable, infectious spores and no signal produced by monolayers inoculated with inactivated spores. Infectivity was expressed as a logit transformation of the proportion of cell culture wells that became infected at each spore challenge dose. The logit dose response method linearizes infectivity data because it excludes 0 percent and 100 percent infectivity and is an accepted method for expressing parasite infectivity. The method did not attempt to use RT-PCR detection in a quantitative manner, but simply as a measure of presence or absence of infection in each of the multiple cell monolayers that were inoculated with each challenge dose of spores. Dose response curves were constructed by fitting a least squares regression to a plot of spore challenge dose against a logit transformation of the proportional infectivity. The ID50 values for individual dose response experiments in RK13 cells ranged from 25 to 252 spores. Based on aggregate data from all experiments with untreated spores, the average ID50 was 36 spores. Following optimization of the cell culture assay, comparable results were obtained for all of the cell lines tested.

Flow cytometric analysis of spores propagated by RK13 cell culture revealed two subpopulations of spores; the larger spores, which are reported to be infectious, accounted for 13 percent of the total population. A possible hypothesis for the decreased infectivity of RK13-propagated E. intestinalis spores that was observed over the course of this research project is that initial spore preparations were comprised of mostly the larger infectious spore forms. With repeated propagation through cell culture, the spore composition shifted so that eventually the majority was the smaller noninfectious form, leading to higher ID50 values. If the smaller spores are immature forms, their presence in cell culture-derived spore populations may indicate that standard cell culture models do not allow all of the parasites to progress to the fully mature spore stage. Increasing host cell differentiation by the addition of butyrate to the medium significantly increased the number of mature spores that could be harvested from the culture. Therefore, polarized cell culture models may be more appropriate than standard nondifferentiated cell lines for propagation of spores for infectivity and disinfection studies.

Human intestinal epithelial cells also were cultured in transwell vessels, allowing cellular polarization, and were used to compare the pathophysiology of infections with E. intestinalis, a human isolate of Brachiola algerae, and Vittaforma corneae. Only E. intestinalis caused significant changes in the permeability of epithelial monolayers, whereas only V. corneae induced apoptosis in infected cells. In long-term infections, E. intestinalis caused loss of brush border and significant swelling. There were substantial differences between the pathophysiology and course of infection associated with these three microsporidia species when assessed in polarized intestinal epithelial cells.

The infectivity assay developed for this research project was used to evaluate the efficacy of UV light for disinfection of E. intestinalis spores. Comparison of the dose response curve for aggregate control experiments with that for UV-exposed spores demonstrated that the average levels of inactivation were 93 percent (1.2-log10) with a UV dose of 3.3 mJ/cm2, 99 percent (2-log10) at 6.2 mJ/cm2, and 99.9 percent (3-log10) at 12.3 mJ/cm2. These results demonstrate that microsporidia spores are inactivated readily by relatively low doses of UV light.

The conclusion from the disinfection work conducted to date and assumptions of physical removal based on size is that if human-pathogenic microsporidia are present in untreated source water, drinking water treatment plants that utilize filtration and disinfection should control them effectively. In light of the heterogeneity of cell culture-derived spore preparations, however, further disinfection studies need to be conducted on only the larger, more infectious spores.

The water industry needs an efficient method for recovering microsporidia from water. This research project, however, has demonstrated the limitations of developing recovery and detection methods for novel microorganisms based on adapting existing techniques. It is apparent that the biological understanding of microsporidia still is evolving, and more information is needed on the morphology of different spore types, nucleic acid sequences, and antigenicity of spores. An efficient recovery procedure will need effective monoclonal antibodies with high specificity for Encephalitozoon spp. and E. bieneusi and strong avidity to their respective targets that can be used for both IMS purification and immunofluorescent detection of spores. Although a fully developed method was not applied to the detection of microsporidia spores in unspiked natural waters, a proposed detection method based on the procedures developed for this research project involves the filtration of 10-100 L samples using 0.55 μm porosity capsule filters or ultrafiltration for larger volumes (100-1,000 L), immunomagnetic purification, quantitative PCR for detection, and cell culture combined with RT-PCR to assess the infectivity of naturally occurring spores. The cell culture-based infectivity assay also can be used to evaluate the efficacy of disinfectants. Based on the disinfection studies conducted to date, including the UV inactivation analyses performed for this research project, human-pathogenic microsporidia should be controlled effectively by drinking water treatment plants that utilize filtration and disinfection.

Journal Articles on this Report : 2 Displayed | Download in RIS Format

Other project views: All 4 publications 2 publications in selected types All 2 journal articles
Type Citation Project Document Sources
Journal Article Kucerova Z, Moura H, Leitch GJ, Sriram R, Bern C, Kawai V, Vargas D, Gilman RH, Ticona E, Vivar A, Visvesvara GS. Purification of Enterocytozoon bieneusi spores from stool specimens by gradient and cell sorting techniques. Journal of Clinical Microbiology 2004;42(7):3256-3261. R828042 (Final)
  • Abstract from PubMed
  • Full-text: ASM Full Text
  • Other: ASM PDF
  • Journal Article Kucerova Z, Moura H, Visvesvara GS, Leitch GJ. Differences between Brachiola (Nosema) algerae isolates of human and insect origin when tested using an in vitro spore germination assay and a cultured cell infection assay. Journal of Eukaryotic Microbiology 2004;51(3):339-343. R828042 (Final)
  • Abstract from PubMed
  • Supplemental Keywords:

    drinking water, human health, pathogens, monitoring, biology, infectivity, microsporidia, spores, disinfection, treatment plants,, RFA, Health, Scientific Discipline, Water, Environmental Chemistry, Health Risk Assessment, Risk Assessments, Environmental Microbiology, Environmental Monitoring, Drinking Water, microbial contamination, pathogens, monitoring, detection, spore germination assay, exposure and effects, exposure, encephalitozoon intestinalis, infectious disease, public health, treatment, microbial risk management, parasites, phase contrast microscopy, infectivity, microsporidia, nosema algare, water treatment, contaminant candidate list, enterocytozoon bieneusi

    Progress and Final Reports:

    Original Abstract
  • 2000