DNA Probe Design for Pre-Flight and In-Flight Microbial Monitoring
George E. Fox, Ph.D., Professor, UH
Richard C. Willson, Ph.D., Associate Professor, UH
Duane L. Pierson, Ph.D., JSC
Karine Maillard, Ph.D., Visiting Assistant Professor and Post-Doctoral Fellow, UH
CREW HEALTH is a dominant issue in manned space flight. Microbiological
concerns, in particular, have repeatedly emerged as determinants of flight readiness. For
example, in at least one case, suspected contamination of the potable water supply nearly
forced a launch delay. In another instance, a crew member's urinary tract infection nearly
led to early termination of the mission, in part due to the difficulty of accurately
diagnosing the nature of the infection in-flight. Microbial problems are an increasing
concern with the trend towards longer-duration missions. It is essential to the success of
such missions that systems be available that deliver acceptable quality of air and water
during the anticipated lifetime of the spacecraft. As mission duration and re-supply
intervals increase, it will be necessary to rely on advanced life support systems which
incorporate both biological and physical-chemical recycling methods for air and water as
well as methodology for providing food for the crew. It is therefore necessary to develop
real-time, robust, in-flight monitoring procedures that are sensitive enough to detect
less than 100 CFU (colony forming units) of bacteria per 100 milliliters of water. It
would be desirable if the monitoring system could be readily "reprogrammed" to
identify specific pathogens if an in-flight incident were to occur. We propose to develop
an appropriate monitoring system based on DNA probe technology.
Above. Dr. George E. Fox, Professor of Biochemical and Biophysical Sciences,
prepares to use a refrigerated sample.
Currently, diagnostics based on nucleic acid hybridization are revolutionizing clinical
identification of difficult bacterial pathogens in terrestrial medicine. Typically, these
tests target ribosomal RNA (rRNA) sequences which are present in all living
organisms. The small subunit rRNA sequence (16S rRNA) has been
experimentally determined in several thousand bacterial species. Each of these sequences
contains short sub-sequences that are widely conserved throughout the data set as well as
other sub-sequences which are totally unique to and, hence, characteristic of a particular
species. This pattern of sequence conservation makes it possible to design oligonucleotide
hybridization probes that can distinguish individual organisms or groupings of related
organisms. Practical diagnostic kits based on this technology (e.g. for Legionella
and Chlamydia), have been successfully introduced into the clinical market. These
single organism diagnostics rely on the natural amplification associated with ribosomal
RNA (as many as 10,000 copies per cell) to provide a probe target and use manual detection
procedures.
Objectives
The long range goal is to develop an automated hybridization system that can be used to
monitor microbial populations and, when necessary, to identify pathogens in air and water
samples during long-duration space missions. The immediate specific aims are: (1) to
design a prototype set of probes that can be used simultaneously to monitor the levels of
total bacteria, fecal coliforms, Escherichia coli, Pseudomonas aeruginosa, Enterococcus
and Burkholderia, (2) to synthesize and test the probes needed to implement the
prototype system using a microtiter-format fluorescence assay. The utility of each probe
will be established individually and then verified when all the probes are used together,
(3) to field-test a ground based implementation of the probe design using a protype
hybridization array, "DNA chip," instrument, and (4) to develop simplified
sample preparation procedures that can be used in the space environment.
Methodology
In order for this technology to be of value in space flight applications, it must
simultaneously detect many organisms of interest, be subject to miniaturization, and be
highly automated. The core test being investigated will use fluorescent or
chemiluminescent deoxyoligonucleotide probes to generate a readily detectable signal. An
especially promising format for automating this basic test is being investigated in
conjunction with Genometrix Inc. and Dr. Michael Hogan at the Baylor College of Medicine.
The essence of the idea is to immobilize an array of unique oligonucleotide probes on a
solid support. Each probe in the array will be designed to be specific for a particular
organism of interest. All probes are simultaneously exposed to the same sample and the
chemiluminescence associated with hybridization is detected by a CCD array. Location of
the hybridization signal on the grid will serve to identify which probes are responsible
for the signal. Since the signal is in discrete locations, the resulting pattern is
subject to straightforward computer analysis. In order to be successful, the approach
requires the availability of sets of probes that can distinguish the organisms of
interest.
A sandwich assay which utilizes both a capture probe and detector probe(s) is being
tested. The surface capture probe selectively acquires the target rRNA from a total
RNA preparation. The detector probe provides a labeling group and assists in the
denaturation of the target molecule by disrupting the secondary structure of the target rRNA
in the region where it binds. The detector probes may provide additional specificity above
that provided by the capture probes or may bind with equal affinity to the 16S rRNAs
of large groups of organisms. Detector probes should not, however, select against the rRNA
of target organisms. Therefore, detector probes must have as much (and preferably more)
sequence identity with the target organisms as they have with any other organism.
Collection of air and water samples in the space environment by filtration is
realistic. However, the methodology to be used for subsequent RNA purification is
uncertain. Purification must be accomplished without the use of either hazardous reagents
or specialized equipment, e.g. centrifugation. UH investigators envision a one- or
two-step column chromatography system as an ideal to simplify materials handling
procedures. Moreover, these methods offer the special advantages of low toxicity, aerosol
generation, and energy consumption.
Progress
Initial hybridization experiments completed in conjunction with Genometrix Inc. and
researchers at the Baylor College of Medicine were performed using pairs of adjacent
homologous/mismatched probes which differentiate E. coli from Vibrio
proteolyticus, but not necessarily from other organisms. Two mismatches between a
probe and a target could be detected using the DNA chip technology. With this experimental
constraint in mind, it is typically possible to identify potential probe target sequences
and capture sequences that are likely to distinguish individual genera or species.
An effective strategy for doing this has been developed. The first step is to obtain
sets of aligned sequences from the public Ribosomal Database Project (RDP) database which
include the target species (or genus) and all closely related organisms. These sequences
are then placed into a sequence editor which allows the sequences to be colorized
according to identity with the target species-and either displayed or printed. It is then
easy to visually identify sequence regions which are likely to distinguish the target
species from other organisms. The next step is to test the putative probe against a
representative subset of the data that includes members from all phylogenetic groups.
Records of numbers of mismatches are compiled at each stage. Finally the most promising
probes are searched against the entire 16S rRNA database for spurious matches.
Surprisingly, more general probes such as an all Bacteria probe and, to a lesser
extent, an enteric probe present a more difficult problem in probe design. This difficulty
arises because there is sufficient variability in the data set as a whole, such that no
one probe of sufficient length will bind to all Bacteria to the exclusion of the Eukaryota
and the Archaea. Our strategy is thus to find two groups of multiple probes specific to
Bacteria, that would be used as capture probes and detector probes: one very specific, and
the other one not as specific.
Due to the large number (now over 4,000) of 16S rRNA sequences in the RDP
database, a computational approach had to be devised to facilitate the search for probes
differentiating Bacteria from Archaea. A PERL program (FIND_ PROBE) based on a
pattern-matching program by Ross Overbeek (PATSCAN) was written. Each time FIND_PROBE is
called, a pattern is chosen from the list of returned patterns, and the corresponding
probe can be easily derived by Watson-Crick base complementarity. In the search for the
all-Bacteria probe, the very specific probes were required to have at least three
mismatches with all Archaea, and the less specific probes were required to have at least
two mismatches with all Archaea. We also set the probe length to be between 13 and 19
nucleotides. The remaining potentially undetectable organisms were sorted
phylogenetically. If it appeared that significant phylogenetic groupings were being
undetected, an additional probe search was performed using an organism from this group.
With this approach, we found three very specific probes, the combination of which would be
expected to detect at least 85 percent of the species of Bacteria in RDP. In most cases,
it was impossible to determine if the remaining species would be detected by this probe
set because one or more of the relevant regions of the 16S rRNA sequence were
undetermined. Analogous approaches were used to design probes for other components of the
proposed preliminary water analysis. In addition to total bacteria, these included probes
for E. coli, total coliforms, Enterococcus, B. cepacia, P.
aeruginosa, and close relatives. Appropriate probes have now been designed for this
purpose. It is next necessary to evaluate experimentally on an individual basis the
specific affinity of each probe for its target RNA. We have therefore implemented sandwich
assay protocols and begun the testing of probes with RNA isolated from the ribosomes of E.
coli, S. xylosus, Bacillus licheniformis, Xanthomonas maltophilia,
B. cepacia, Methylobacterium sp., and Halobacterium vallismortui.
In order to demonstrate that designed probes are appropriate for organism strains
actually encountered in the space environment as well as for standard database strains, we
undertook the sequencing of 16S rRNAs from several bacterial strains that were
isolated from water samples taken on the Mir station. These organisms had previously been
identified by conventional methods as B. licheniformis, X. maltophilia, B.
cepacia, and Methylobacterium sp. The results confirmed the microbiological
identification of the species and showed that the expected probe target sequences were
indeed in these organisms.
In order to facilitate sample preparation, we are investigating the possible utility of
adsorptive separations as the basis of spacecraft-compatible methods of sample preparation
for DNA probe analysis. We have confirmed that immobilized metals capture RNA effectively,
with selectivity over DNA. Immobilized metal-chelating functionalities charged with Cu++
are able to capture RNA species from solution, presumably by interaction with the nitrogen
heterocyclic bases, which are relatively exposed in the single-stranded portions of RNA.
This interaction appears to be mediated by base nitrogen heterocycles. Thus, immobilized
metals, chelated by NTA or IDA ligands, may be promising as semi-selective capture agents
for the concentration and partial purification of rRNA. As an efficient
cross-calibration technique for probe development and testing, we have also developed
quantitative probe hybridization assays using a phosphorimager, a device that gives rapid,
reproducible results, even with fairly complex starting samples.
 |
| UH Investigators Dr. Karine Maillard (l.) with Dr. Richard
Willson conducting research in DNA probe design and in-flight microbial monitoring. |
Contents
ISSO -- Institute for Space Systems Operations
1996-1997 Annual Report
