Chapter 3: Classification of Bacteria

One has only to peruse the table of contents of this book to appreciate the diversity of medical pathogens that are associated with infectious diseases. It has been estimated that we currently have the capacity to identify fewer than 10% of the pathogens responsible for causing human disease. This is due to our inability to culture or target these organisms using molecular probes. The diversity of even these identifiable pathogens alone is so great that it is important to appreciate the subtleties associated with each infectious agent. The reason for understanding these differences is significant because each infectious agent has specifically adapted to a particular mode(s) of transmission, a mechanism(s) to grow in a human host (colonization), and a mechanism(s) to cause disease (pathology). As such, a vocabulary that consistently communicates the unique characteristics of infectious organisms to students, microbiologists, and health care workers is critical to avoid the chaos that would ensue without the organizational guidelines of bacterial taxonomy (Gk. taxon = arrangement; eg, the classification of organisms in an ordered system that indicates a natural relationship).

Identification, classification, and nomenclature are three separate but interrelated areas of bacterial taxonomy. Each area is critical to the ultimate goal of accurately studying the infectious diseases and precisely communicating these to others in the field.

Identification is practical use of a classification scheme to (1) isolate and distinguish specific organisms among the mix of complex microbial flora, (2) verify the authenticity or special properties of a culture in a clinical setting, and (3) isolate the causative agent of a disease. The latter may lead to the selection of specific pharmacologic treatments directed toward their eradication, a vaccine mitigating their pathology, or a public health measure (eg, handwashing) that prevents further transmission.

Identification schemes are not classification schemes, although there may be some superficial similarity. For example, the popular literature has reported Escherichia coli as being a cause of hemolytic uremic syndrome (HUS) in infants. There are hundreds of different strains that are classified as E coli but only a few that are associated with HUS. These strains can be “identified” from the many other E coli strains by antibody reactivity with their O-, H-, and K-antigens, as described in Chapter 2 (eg, E coli O157:H7). However, they are more broadly classified as a member of the family Enterobacteriaceae.

In a microbiologic context, classification is the categorization of organisms into taxonomic groups. Experimental and observational techniques are required for taxonomic classification. This is because biochemical, physiologic, genetic, and morphologic properties are historically necessary for establishing a taxonomic rank. This area of microbiology is necessarily dynamic as the tools continue to evolve (eg, new methods of microscopy, biochemical analysis, and computational nucleic acid biology).

Nomenclature refers to the naming of an organism by an established group of scientific and medical professionals. This is arguably the most important component of taxonomy because it allows medical professionals to communicate with each. Just as our societal vocabulary evolves, so does the vocabulary of medical microbiology. Any professional associated with infectious disease should be aware of the evolving taxonomy of infectious microorganisms.

Ultimately, the taxonomic ranks form the basis for the organization of bacteria. Linnaean taxonomy is the system most familiar to biologists. It uses the formal ranks of kingdom, phylum, class, order, family, genus, and species. The lower ranks are approved by a consensus of experts in the scientific community. Of these ranks, the family, genus, and species are the most useful (Table 3-1).

Growth on Media

Suitable criteria for bacterial classification include many of the properties that were described in the preceding chapter. One criterion is growth on different types of bacteriologic media. The general cultivation of most bacteria requires media rich in metabolic nutrients. These media generally include agar, a carbon source, and an acid hydrolysate or enzymatically degraded source of biologic material (eg, casein). Because of the undefined composition of the latter, these types of media are referred to as complex media.

Clinical samples from normally nonsterile sites (eg, the throat or the colon) contain multiple species of organisms, including potential pathogens and resident microbial flora. Media can be nonselective or selective; the latter are used to distinguish among the various bacteria in a clinical sample containing many different organisms.

A. Nonselective Media

Blood agar and chocolate agar are examples of complex, nonselective media, which support the growth of many different bacteria. These media are intended to cultivate as many species as possible, thus giving rise to numerous types of bacterial colonies.

B. Selective Media

Because of the diversity of microorganisms that typically reside at some sampling sites (eg, the skin, respiratory tract, intestines, vagina), selective media are used to eliminate (or reduce) the large numbers of irrelevant bacteria in these specimens. The basis for selective media is the incorporation of an inhibitory agent that specifically selects against the growth of irrelevant bacteria. Examples of such agents are:

• Sodium azide—selects for gram-positive bacteria over gram-negative bacteria

• Bile salts (sodium deoxycholate)—select for gram-negative enteric bacteria and inhibit gram-negative mucosal and most gram-positive bacteria

• Colistin and nalidixic acid—inhibit the growth of many gram-negative bacteria

Examples of selective media are MacConkey agar (contains bile) that selects for the Enterobacteriaceae and CNA blood agar (contains colistin and nalidixic acid) that selects for staphylococci and streptococci.

C. Differential Media

Upon culture, some bacteria produce characteristic pigments, and others can be differentiated on the basis of their complement of extracellular enzymes; the activity of these enzymes often can be detected as zones of clearing surrounding colonies grown in the presence of insoluble substrates (eg, zones of hemolysis in agar medium containing red blood cells).

Many of the members of the Enterobacteriaceae can be differentiated on the basis of their ability to metabolize lactose. For example, pathogenic salmonellae and shigellae that do not ferment lactose on a MacConkey plate form white colonies, while lactose-fermenting members of the Enterobacteriaceae (eg, E coli) form red or pink colonies.

The number of differential media used in today’s clinical laboratories is far beyond the scope of this chapter.

Microscopy

Historically, the Gram stain, together with visualization by light microscopy, has been among the most informative methods for classifying the eubacteria. This staining technique broadly divides bacteria on the basis of fundamental differences in the structure of their cell walls (see Chapter 2). This typically represents the first step in identifying individual microbial specimens (eg, are they gram negative or gram positive) grown in culture or even directly from patient specimens (eg, urine specimens).

Biochemical Tests

Tests such as the oxidase test, which uses an artificial electron acceptor, can be used to distinguish organisms on the basis of the presence or absence of a respiratory enzyme, cytochrome C, the lack of which differentiates the Enterobacteriaceae from other gram-negative rods. Similarly, catalase activity can be used, for example, to differentiate between the gram-positive cocci; the species staphylococci are catalase positive, whereas the species streptococci are catalase negative. If the organism is demonstrated to be catalase positive (Staphylococcus spp.), the species can be subdivided by a coagulase test into Staphylococcus aureus (coagulase positive) or Staphylococcus epidermitidis (coagulase negative) as demonstrated in Figure 3-1.

Ultimately, there are many examples of biochemical tests that can ascertain the presence of characteristic metabolic functions and be used to group bacteria into a specific taxon. A nonexhaustive list of common biochemical tests is given in Table 3-2.

Immunologic Tests—Serotypes, Serogroups, and Serovars

The designation “sero” simply indicates the use of antibodies (polyclonal or monoclonal) that react with specific bacterial cell surface structures such as lipopolysaccharide (LPS), flagella, or capsular antigens. The terms “serotype,” “serogroups,” and “serovars” are, for all practical purposes, identical—they all use the specificity of these antibodies to subdivide strains of a particular bacterial species.

Under certain circumstances (eg, an epidemic), it is important to distinguish among strains of a given species or to identify a particular strain. This is called subtyping and is accomplished by examining bacterial isolates for characteristics that allow discrimination below the species level. Classically, subtyping has been accomplished by biotyping, serotyping, antimicrobial susceptibility testing, and bacteriophage typing. For example, more than 130 serogroups of Vibrio cholerae have been identified on the basis of antigenic differences in the O-polysaccharide of their LPS; however, only the O1 and O139 serogroups are associated with epidemic and pandemic cholera. Within these serogroups, only strains that produce a particular toxin-coregulated pili and cholera toxin are virulent and cause the disease cholera. By contrast, nontoxigenic V cholerae strains, which are not associated with epidemic cholera, have been isolated from environmental specimens, from food, and from patients with sporadic diarrhea.

Clonality with respect to isolates of microorganisms from a common source outbreak (point source spread) is an important concept in the epidemiology of infectious diseases. Etiologic agents associated with these outbreaks are generally clonal; in other words, they are the progeny of a single cell and thus, for all practical purposes, are genetically identical. Thus, subtyping plays an important role in discriminating among these particular microorganisms. Recent advances in biotechnology have dramatically improved our ability to subtype microorganisms. Hybridoma technology has resulted in the development of monoclonal antibodies against cell surface antigens, which have been used to create highly standardized antibody-based subtyping systems that describe bacterial serotypes. This is an important tool for defining the epidemiologic spread of a bacterial infection.

Other organisms cannot be identified as unique serotypes. For example, some pathogens (eg, Neisseria gonorrhoeae) are transmitted as an inoculum composed of quasispecies (meaning that there is extensive antigenic variation among the bacteria present in the inoculum). In these cases, groups of hybridomas that recognize variants of the original organisms are used to categorize serovariants or serovars.

Genetic Diversity

The value of a taxonomic criterion depends upon the biologic group being compared. Traits shared by all or none of the members of a group cannot be used to distinguish its members, but they may define a group (eg, all staphylococci produce the enzyme catalase). Developments in molecular biology now make it possible to investigate the relatedness of genes or genomes by comparing sequences among different bacteria. It should be noted that genetic instability can cause some traits to be highly variable within a biologic group or even within a specific taxonomic group. For example, antibiotic resistance genes or genes encoding enzymes (eg, lactose utilization) may be carried on plasmids or bacteriophages (see Chapter 7), extrachromosomal genetic elements that may be transferred among unrelated bacteria or that may be lost from a subset of bacterial strains identical in all other respects. Many organisms are difficult to cultivate, and in these instances, techniques that reveal relatedness by measurement of nucleic acid hybridization or by DNA sequence analysis may be of particular value.

Dichotomous Keys

Dichotomous keys organize bacterial traits in a manner that permits logical identification of organisms. The ideal system should contain the minimum number of features required for a correct categorization. Groups are split into smaller subgroups based on the presence (+) or absence (−) of a diagnostic character. Continuation of the process with different characteristics guides the investigator to the smallest defined subgroup containing the analyzed organism. In the early stages of this process, organisms may be assigned to subgroups on the basis of characteristics that do not reflect genetic relatedness. It would be perfectly reasonable, for example, for a bacterial key to include a group such as “bacteria forming red pigments when propagated on a defined medium” even though this would include such unrelated forms as Serratia marcescens (see Chapter 15) and purple photosynthetic bacteria (see Chapter 6). These two disparate bacterial assemblages occupy distinct niches and depend on entirely different forms of energy metabolism. Nevertheless, preliminary grouping of the assemblages would be useful because it would immediately make it possible for an investigator having to identify a red-pigmented culture to narrow the range of possibilities to relatively few groups. An example of a dichotomous key is shown in Figure 3-1.

Numerical Taxonomy Using Biochemical Measures of Activity

Numerical taxonomy became widely used in the 1970s. These classification schemes use a large number of unweighted, taxonomically useful characteristics. For these assays, an individual bacterial colony must be isolated and used to inoculate the test format. One example of this is the Analytical Profile Index (API^TM), which uses numerical taxonomy to identify a wide range of medically important microorganisms. APIs consist of a number of plastic strips, each of which has about 20 miniature compartments containing biochemical reagents (Figure 3-2). Almost all cultivatable bacterial groups and more than 550 different species can be identified using the results of these API tests. These identification systems have extensive databases of microbial biochemical reactions. The numerical clusters derived from these tests identify different strains at selected levels of overall similarity (usually >80% at the species level) on the basis of the frequency with which they share traits. In addition, numerical classification provides percentage frequencies of positive character states for all strains within each cluster. The limitation of this approach is that it is a static system. As such, it does not allow for the evolution of bacteria and routine discovery of new bacterial pathogens.

Nucleic Acid–Based Taxonomy

Since 1975, developments in nucleic acid isolation, amplification, and sequencing spurred the evolution of nucleic acid–based subtyping systems. These include plasmid profile analysis, restriction fragment endonuclease analysis, repetitive sequence analysis, ribotyping, and genomic sequencing. These methods are individually described as follows.

Plasmid Analysis

Plasmids are extrachromosomal genetic elements (see Chapter 7). These can be isolated from an isolated bacterium and separated by agarose gel electrophoresis to determine their number and size. Plasmid analysis has been shown to be most useful for examining outbreaks that are restricted in time and place (eg, an outbreak in a hospital), particularly when they are combined with other identification methods.

Restriction Endonuclease Analysis

The use of restriction enzymes to cleave DNA into discrete fragments is one of the most basic procedures in molecular biology. Restriction endonucleases recognize short DNA sequences (restriction sequence), and they cleave double-stranded DNA within or adjacent to this sequence. Restriction sequences range from 4 to more than 12 bases in length and occur throughout the bacterial chromosome. Restriction enzymes that recognize short sequences (eg, 4-base pairs) occur more frequently than those that are specific for longer sequences (eg, 12-base pairs). Thus, enzymes that recognize short DNA sequences produce more fragments than enzymes that recognize infrequently occurring long DNA sequences. Several subtyping methods use restriction endonuclease–digested DNA.

One method involves isolating the plasmid DNA, which is generally of the size of several kilobases, and digesting this nucleic acid with a restriction enzyme. After enzymatic cleavage, fragmented plasmid segments are separated using agarose gel electrophoresis. Because plasmids carry genetic material that directly contribute to disease and are commonly moved from one organism to another, the presence of a common fragment may confirm that a specific bacterial isolate was identical to other isolates associated with an outbreak.

Another method involves the analysis of genomic DNA, which is of the size of several megabases. In this case, restriction endonucleases that cut at infrequently occurring restriction sites within the bacterial genome are used. Digestion of DNA with these enzymes generally results in 5–20 fragments ranging from approximately 10 to 800 kb in length. Separation of these large DNA fragments is accomplished by a technique called pulsed field gel electrophoresis (PFGE), which requires specialized equipment. Theoretically, all bacterial isolates can be typed by this method. Its advantage is that the restriction profile consists of a finite number of well-resolved bands representing the entire bacterial chromosome in a single DNA fragment pattern.

Genomic Analysis

The routine use of DNA genome sequencing allows the precise comparison of divergent DNA sequences, which can give a measure of their relatedness. Genes for different functions, such as those encoding surface antigens to escape immune surveillance, diverge at different rates relative to “housekeeping” genes such as those that encode cytochromes. Thus, DNA sequence differences among rapidly diverging genes can be used to ascertain the genetic distance between closely related groups of bacteria. Sequence differences among housekeeping genes can be used to measure the relatedness of widely divergent groups of bacteria.

The genetic properties of bacteria allow genes to be exchanged among distantly related organisms. Furthermore, multiplication of bacteria is almost entirely vegetative, and their mechanisms of genetic exchange rarely involve recombination among large portions of their genomes (see Chapter 7). Therefore, the concept of a species—the fundamental unit of eukaryotic phylogeny—has an entirely different meaning when applied to bacteria.

There is considerable genetic diversity among bacterial species. Chemical characterization of bacterial genomic DNA reveals a wide range of nucleotide base compositions among different bacterial species. One measure of this is the guanine + cytosine (G + C) content. If the G + C content of two different bacterial species is similar, it indicates taxonomic relatedness.

Repetitive Sequence Analysis

In the current genomic era of molecular medicine, hundreds of microbial genomes have now been sequenced. With this era have come bioinformatical tools to mine this wealth of DNA sequence information to identify novel targets for pathogen subtyping, such as the repetitive sequences that have been found in different species (see Chapter 7). These repetitive sequences have been termed satellite DNA and have repeating units that range from 10 to 100 bp. They are commonly referred to as variable number tandem repeats (VNTRs). VNTRs have been found in regions controlling gene expression and within open reading frames. The repeat unit and the number of copies repeated side by side define each VNTR locus. A genotyping approach using polymerase chain reaction (PCR), referred to as multiple-locus VNTR analysis (MLVA), takes advantage of the levels of diversity generated by both repeat unit size variation and copy number among a number of characterized loci. It has proved especially useful in subtyping monomorphic species such as Bacillus anthracis, Yersinia pestis, and Francisella tularensis.

Ribosomal RNA

Ribosomes have an essential role in protein synthesis for all organisms. Genetic sequence encodings both ribosomal RNAs (rRNA) and proteins (both of which are required to comprise a functional ribosome) have been highly conserved throughout evolution and have diverged more slowly than other chromosomal genes. Comparison of the nucleotide sequence of 16S rRNA from a range of prokaryotic sources revealed evolutionary relationships among widely divergent organisms and has led to the elucidation of a new kingdom, the archaebacteria. The phylogenetic tree based on rRNA data, showing the separation of bacteria, archaea, and eukaryote families, is depicted in Figure 3-3, which shows the three major domains of biological life as they are currently understood. From this diagram, two kingdoms, the eubacteria (true bacteria) and the archaebacteria, are distinct from the Eukaryotic branch.

The technique of Southern blot analysis was named after its inventor, Edwin Mellor Southern, and has been used as a subtyping method to identify isolates associated with outbreaks. For this analysis, DNA preparations from bacterial isolates are subjected to restriction endonuclease digestion. After agarose gel electrophoresis, the separated restriction fragments are transferred to a nitrocellulose or nylon membrane. These double-stranded DNA fragments are first converted into single-stranded linear sequences. Using a labeled fragment of DNA as a probe, it is possible to identify the restriction fragments containing sequences (loci) that are homologous to the probe by complementation to the bound single-stranded fragments (Figure 3-4).

Southern blot analysis can be used to detect polymorphisms of rRNA genes, which are present in all bacteria. Because ribosomal sequences are highly conserved, they can be detected with a common probe prepared from the 16S and 23S rRNA of a eubacterium (see Figure 3-6). Many organisms have multiple copies (five to seven) of these genes, resulting in patterns with a sufficient number of bands to provide good discriminatory power; however, ribotyping are of limited value for some microorganisms, such as mycobacteria, which have only a single copy of these genes.

Bergey’s Manual of Systematic Bacteriology

The definitive work on the taxonomic organization of bacteria is the latest edition of Bergey’s Manual of Systematic Bacteriology. First published in 1923, this publication taxonomically classifies, in the form of a key, known bacteria that have or have not been cultured or well described. A companion volume, Bergey’s Manual of Determinative Bacteriology, serves as an aid in the identification of bacteria that have been described and cultured. The major bacteria that cause infectious diseases, as categorized in Bergey’s Manual, are listed in Table 3-3. Because it is likely that emerging information concerning phylogenetic relationships will lead to further modifications in the organization of bacterial groups within Bergey’s Manual, its designations must be regarded as a work in progress.

As discussed in Chapter 2, there are two different groups of prokaryotic organisms, eubacteria and archaebacteria. Both are small unicellular organisms that replicate asexually. Eubacteria refer to classic bacteria as science has historically understood them. They lack a true nucleus, have characteristic lipids that make up their membranes, possess a peptidoglycan cell wall, and have a protein and nucleic acid synthesis machinery that can be selectively inhibited by antimicrobial agents. In contrast, archaebacteria do not have a classic peptidoglycan cell wall and have many characteristics (eg, protein synthesis and nucleic acid replication machinery) that are similar to those of eukaryotic cells.

The Eubacteria

A. Gram-Negative Eubacteria

This is a heterogeneous group of bacteria that have a complex (gram-negative type) cell envelope consisting of an outer membrane, a periplasmic space containing a thin peptidoglycan layer, and a cytoplasmic membrane. The cell shape (Figure 3-5) may be spherical, oval, straight or curved rods, helical, or filamentous; some of these forms may be sheathed or encapsulated. Reproduction is by binary fission, but some groups reproduce by budding. Fruiting bodies and myxospores may be formed by the myxobacteria. Motility, if present, occurs by means of flagella or by gliding motility. Members of this category may be phototrophic or nonphototrophic (see Chapter 5) bacteria and include aerobic, anaerobic, facultatively anaerobic, and microaerophilic species.

B. Gram-Positive Eubacteria

These bacteria have a cell wall profile of the gram-positive type; cells generally, but not always, stain gram positive. The cell envelope of gram-positive organisms consists of a thick cell wall that determines cellular shape and a cytoplasmic membrane. These cells may be encapsulated and can exhibit flagellar-mediated motility. Cells may be spherical, rods, or filaments; the rods and filaments may be nonbranching or may show true branching. Reproduction is generally by binary fission. Some bacteria in this category produce spores (eg, Bacillus and Clostridium spp.) as resting forms that are highly resistant to disinfection. The gram-positive eubacteria are generally chemosynthetic heterotrophs (see Chapter 5) and include aerobic, anaerobic, and facultatively anaerobic species. The groups within this category include simple asporogenous and sporogenous bacteria as well as the structurally complex actinomycetes and their relatives.

C. Eubacteria Lacking Cell Walls

These are microorganisms that lack cell walls (commonly called mycoplasmas and making up the class Mollicutes) and do not synthesize the precursors of peptidoglycan. They are enclosed by a unit membrane, the plasma membrane (Figure 3-6). They resemble the L-forms that can be generated from many species of bacteria (notably gram-positive eubacteria); unlike L-forms, however, mycoplasmas never revert to the walled state, and there are no antigenic relationships between mycoplasmas and eubacterial L-forms.

Six genera have been designated as mycoplasmas on the basis of their habitat; however, only two genera contain animal pathogens. Mycoplasmas are highly pleomorphic organisms and range in size from vesicle-like forms to very small (0.2 μm), filterable forms (meaning that they are too small to be captured on filters that routinely trap most bacteria). Reproduction may be by budding, fragmentation, or binary fission, singly or in combination. Most species require a complex medium for growth and tend to form characteristic “fried egg” colonies on a solid medium. A unique characteristic of the Mollicutes is that some genera require cholesterol for growth; unesterified cholesterol is a unique component of the membranes of both sterol-requiring and non–sterol-requiring species if present in the medium.

The Archaebacteria

These organisms are predominantly inhabitants of extreme terrestrial and aquatic environments (high salt, high temperature, anaerobic) and are often referred to as extremeophiles; some are symbionts in the digestive tract of humans and animals. The archaebacteria consist of aerobic, anaerobic, and facultatively anaerobic organisms that are chemolithotrophs, heterotrophs, or facultative heterotrophs. Some species are mesophiles, but others are capable of growing at temperatures above 100°C. These hyperthermophilic archaebacteria are uniquely adapted for growth at high temperatures. With few exceptions, enzymes isolated from these organisms are intrinsically more thermostable than their counterparts from mesophilic organisms. Some of these thermostable enzymes, such as the DNA polymerase from Thermus aquaticus (Taq polymerase), are important components of DNA amplification methods such as the PCR.

Archaebacteria can be distinguished from eubacteria in part by their lack of a peptidoglycan cell wall, possession of isoprenoid diether or diglycerol tetraether lipids, and characteristic rRNA sequences. Archaebacteria also share some molecular features with eubacteria (Table 3-4). Cells may have a diversity of shapes, including spherical, spiral, and plate or rod shaped; unicellular and multicellular forms in filaments or aggregates also occur. Multiplication occurs by binary fission, budding, constriction, fragmentation, or other unknown mechanisms.

Attempts to estimate total numbers of eubacteria and archaebacteria are problematic because of difficulties such as detection in and recovery from the environment. As indicated earlier, estimates indicate that the numbers of unculturable microbial taxa greatly exceed those of the culturable organisms. Recent estimates suggest that the number of bacterial species in the world is between 10⁷ and 10⁹. Until very recently, microbial identification required the isolation of pure cultures followed by testing for multiple physiologic and biochemical traits. Clinicians have long been aware of human diseases that are associated with visible but nonculturable microorganisms. Scientists are now using a PCR-assisted approach using rRNA to identify pathogenic microorganisms in situ. The first phase of this approach involves the extraction of DNA from a suitable specimen, the use of standard molecular techniques to obtain a clone library, the retrieval of rRNA sequence information, and a comparative analysis of the retrieved sequences. This yields information on the identity or relatedness of the sequences in comparison with the available database. In the second phase, proof that the sequences are from cells in the original specimen is obtained by in situ hybridization using sequence-specific probes. This approach has been used in the identification of pathogenic microorganisms. For example, a previously uncharacterized pathogen has been identified as the Whipple-disease–associated rod-shaped bacterium now designated Tropheryma whipplei. This rRNA approach has also been used to identify the etiologic agent of bacillary angiomatosis as Bartonella henselae and to show that the opportunistic pathogen Pneumocystis jiroveci is a member of the fungi. Undoubtedly, these and other techniques will identify additional etiologic agents in the future.

1. Understand how the vocabulary of taxonomy is critical to communicating the science of infectious diseases.

2. Know the taxonomic ranks.

3. Appreciate the growth, biochemical, and genetic characteristics that are used in differentiating bacteria.

4. Understand the differences between the eubacteria, archaebacteria, and eukaryotes.

5. Be conversant in the different tools for nucleic acid–based taxonomy.