| Bacterial identification in
the diagnostic laboratory
Isolation and identification of
bacteria from patients aids treatment since infectious diseases caused
by different bacteria have a variety of clinical courses and
consequences. Susceptibility testing of isolates (i.e. establishing the
minimal inhibitory concentration or MIC) can help in selection of
antibiotics for therapy. Recognizing that certain species (or strains)
are being isolated atypically may suggest that an outbreak has occurred
e.g. from contaminated hospital supplies or poor aseptic technique on
the part of hospital personnel.
When patients are suspected of
having a bacterial infection, it is usual to isolate visible colonies of
the organism in pure culture (on agar plates), and then speciate the
organism. The identification is based on taxonomic principles applied to
the clinical microbiological situation. In the diagnostic laboratory,
many samples must be characterized each day and results obtained as
quickly as possible. Tests must be easily learned, low in cost and
rapidly performed. These classical methods for speciation of bacteria
are based on morphological and metabolic characteristics. The diagnostic
tests have been selected on the basis that empirically they provide
discriminating information. There are numerous different tests for each
of the many target pathogens. Additionally, molecular biology techniques
(for characterization of specific genes or gene segments) are now
commonplace in the clinical laboratory.
Bicarbonate and blood agar plate cultures of Bacillus anthracis.
Smooth colonies on bicarbonate (left) and rough colonies on blood
agar (right). CDC/Dr. James Feeley
Modern taxonomic approaches often
employ technically more complex methodology and are concerned with
profiling the structural composition of bacteria. This often involves
"molecular biology" or "analytical chemical" based
approaches. It is now recognized that many of the classical schemes for
differentiation of bacteria provide little insight into their genetic
relationships and in some instances are scientifically incorrect. New
information has resulted in renaming of certain bacterial species and in
some instances has required totally reorganizing relationships within
and between many bacterial families.
Family: a group of related
Genus: a group of related
Species: a group of related
Type: sets of strain within a
species (e.g. biotypes, serotypes).
Strain: one line or a single
isolate of a particular species.
The most commonly used term is
the species name (e.g. Streptococcus pyogenes or Streptococcus
pyogenes - abbreviation S.pyogenes). There is always two
parts to the species name, one defining the genus in this case "Streptococcus"
and the other the species (in this case "pyogenes"). The genus
name is always capitalized but the species name is not. Both species and
genus are underlined or in italics.
Steps in diagnostic isolation
and identification of bacteria
Step 1. Samples of body fluids
(e.g. blood, urine, cerebrospinal fluid) are streaked on culture plates
and isolated colonies of bacteria (which are visible to the naked eye)
appear after incubation for one - several days. Each colony consists of
millions of bacterial cells. Observation of these colonies for size,
texture, color, and (if grown on blood agar) hemolysis reactions, is
highly important as a first step in bacterial identification. Whether
the organism requires oxygen for growth is another important
Step 2. Colonies are Gram stained
and individual bacterial cells observed under the microscope.
Step 3. The bacteria are
speciated using these isolated colonies. This often requires an
additional 24 hr of growth.
THE GRAM STAIN
A colony is dried on a slide and
treated as follows:
Step 1. Staining with crystal
Step 2. Fixation with iodine
stabilizes crystal violet staining. All bacteria remain purple or blue.
Step 3. Extraction with alcohol
or other solvent. Decolorizes some bacteria (Gram negative) and not
others (Gram positive).
Step 4. Counterstaining with
safranin. Gram positive bacteria are already stained with crystal violet
and remain purple. Gram negative bacteria are stained pink.
Under the microscope, the
appearance of bacteria is observed. Questions to be asked include: Are
they Gram positive or negative? What is the morphology (rod, coccus,
spiral, pleomorphic [variable form] etc)? Do cells occur singly or in
chains, pairs etc? How large are the cells? There are other less
commonly employed stains available (e.g. for spores and capsules).
Another similar colony from the
primary isolation plate is then examined for biochemical properties
(e.g. will it ferment a sugar such as lactose). In some instances the
bacteria are identified (e.g. by aggregation) with commercially
available antibodies recognizing defined surface antigens. Molecular
tests are now widely used.
There is considerable diversity
even within a species. Thus comparisons of species involve comparisons
of multiple strains for each species. Comparisons are primarily based on
chemical or molecular analysis.
Sophisticated tools are available for studying the structural
composition of bacteria (most commonly fatty acid, carbohydrate or ubiquinone
profiling) of bacteria. Characterization of secreted metabolic products
(e.g. volatile alcohols and short chain fatty acids) is also helpful.
Molecular analysis It
would be ideal to compare sequences of entire bacterial chromosomal DNA,
but this is currently not feasible. Millions of nucleotides have to be
sequenced for each strain. In the past several years, sequencing of the
entire genomes of one representative (i.e. a strain) of a few bacterial
species has been achieved. In each case, this has involved massive
amounts of work by large research groups dedicated to the task of
sequencing. Alternatively, genomic similarity has been historically
assessed by the content of guanine (G) + cytosine (C), usually expressed
as a percentage (% GC). This has been replaced by two alternatives -
hybridization and sequencing (most commonly of the gene coding for 16S
DNA-DNA homology (or how well two
strands of DNA from different bacteria bind [hybridize] together) is
employed to compare the genetic relatedness of bacterial
strains/species. If the DNA from two bacterial strains display a high
degree of homology (i.e. they bind well) the strains are considered to
be members of the same species. DNA from different bacterial species
(unless closely related) display no homology.
In the last few years sequencing
of 16S ribosomal RNA molecules (16S rRNA) has become the "gold
standard" in bacterial taxonomy. The molecule is approximately
sixteen hundred nucleotides in length. The sequence of 16S rRNA provides
a measure of genomic similarity above the level of the species allowing
comparisons of relatedness across the entire bacterial kingdom. Closely
related bacterial species often have identical rRNA sequences. The
technique thus provides complementary information to DNA-DNA
hybridization. Exploiting this basic research in the clinical laboratory
has allowed the development of probes for improved identification of
Approaches to rapid diagnosis
without prior culture
Certain human pathogens
(including the causative agents of tuberculosis, Lyme disease and
syphilis) either cannot be isolated in the laboratory or grow extremely
poorly. Successful isolation can be slow and in some instances
impossible. Direct detection of bacteria without culture is possible in
A simple approach to rapid
diagnosis (as an example of antigen detection) is used in many doctor's
offices for the group A streptococcus. The patient's throat is swabbed
and streptococcal antigen extracted directly from the swab (without
prior bacteriological culture). The bacterial antigen is detected by
aggregation (agglutination) of antibody coated latex beads.
Bacterial DNA sequences can be
amplified directly from human body fluids (the polymerase chain
reaction, PCR). In this fashion large amounts of specific genes or
portions of genes can be generated and readily detected. For example,
great success has been achieved in rapid diagnosis of tuberculosis.
Finally, direct microscopic
observation of certain clinical samples for the presence of bacteria can
be helpful (e.g. detection of M. tuberculosis in sputum).
Serologic identification of an
antibody response (in patient's serum) to the infecting agent can only
be successful several weeks after an infection has occurred.