The canine and feline skin microbiome in health and disease
The skin harbours a diverse and abundant, yet inadequately investigated, microbial population. The population is believed to play an important role in both the pathophysiology and the prevention of disease, through a variety of poorly explored mechanisms.
Early studies of the skin microbiota in dogs and cats reported a minimally diverse microbial composition of low overall abundance, most probably as a reflection of the limitations of testing methodology. Despite these limitations, it was clear that the bacterial population of the skin plays an important role in disease and in changes in response to both infectious and noninfectious diseases.
Recent advances in technology are challenging some previous assumptions about the canine and feline skin microbiota and, with preliminary application of next-generation sequenced-based methods, it is apparent that the diversity and complexity of the canine skin microbiome has been greatly underestimated. A better understanding of this complex microbial population is critical for elucidation of the pathophysiology of various dermatological (and perhaps systemic) diseases and to develop novel ways to manipulate this microbial population to prevent or treat disease.
In many ways, the skin represents an inhospitable site for micro-organisms, as a relatively cool, dry, high-salt, hydrophobic and acidic environment that is covered with antibacterial peptides and constantly shedding superficial layers.1 Yet, despite these challenges, the skin harbours a large and complex microbial population, with estimates of 1 million to 1 billion bacteria per square centimetre of skin in humans.2,3 This population of micro-organisms (the microbiota), particularly the bacterial component, has an intimate relationship with the host and plays a role in both protection and development of disease. Despite the recognition of the importance of the skin microbiota and the sum of its genetic components (the microbiome), understanding of this complex microbial environment is superficial, and approaches to modifying it (e.g. antimicrobials) have been rather crude.
Assessment of microbial populations
Understanding the structure and function of a microbial population requires methods to study the individual components of that population, something that can be challenging with large, complex microbial ecosystems. Traditionally, bacterial and fungal populations are assessed through conventional culture methods. While useful, culture has significant limitations in determining the overall microbiota because of the variable ability to culture different bacteria and limitations in the depth of study, which hamper investigation of locations that may contain millions to billions of bacteria from a variety of diverse genera. Even organisms that grow using conventional methods may be missed or underestimated when they are part of a complex population that includes other bacteria that grow readily in standard conditions (e.g. staphylococci). Accordingly, culture-based studies inherently lead to underestimation of the diversity and abundance, and over-estimation of the presence of certain species or groups.2
Recognition of limitations in conventional culture and the need for methods that provide more depth of study led to development of culture-independent methods. These have been best explored in the gastrointestinal tract, and the shift towards nonculture-dependent methodologies has led to a revolution in our understanding of the body’s microbial composition. A variety of culture-independent methods are available, each with advantages and disadvantages,4,5 but most efforts are now focused on the use of high-throughput next-generation sequencing methods that are able to differentiate large numbers of individual sequences from complex mixtures. While still prone to some biases, particularly if inadequate quality control and data cleaning methods are used, this approach is providing tremendous new insight into our knowledge of microbiomes.
The human skin
The human skin has been the most intensively studied skin microbiome, albeit with a fraction of the effort dedicated to the gastrointestinal tract. Classically, the skin microbiota is divided into two groups, the resident flora and the transient flora. The resident flora comprises the core microbiota that is relatively consistent and stable, and which repopulates itself rapidly after disruption.6,7 This group is considered the true commensal microbiota. In contrast, the transient flora is an ever-changing population of temporary inhabitants that arise from the individual’s environment or other external contacts and only persist on the skin for hours to days.2,6 Properly defining resident and transient populations is dependent on the quality of the laboratory technique used, and therefore current information is at best incomplete, but the concept of a common, abundant and stable population living with a transient, dynamic and variable population is likely to be valid. Both groups can contain a range of harmless commensals and potential pathogens, yet disease occurs uncommonly despite the relative abundance of many well-adapted opportunistic pathogens. Given the proper circumstances, pathogenic members of both the resident and the transient microbiota can cause disease, and status as a member of the transient or resident microbiota is not necessarily an indication of the virulence of a micro-organism.
The bacterial microbiota: culture-based investigation
Early culture-based studies of the skin microbiota provided what is known to be rather superficial, yet critical, information.8,9 Staphylococci were identified as predominant components of human skin, particularly Staphylococcus epidermidis and other coagulase-negative species. Staphylococcus aureus was recognized as an important cause of skin infection but also a relatively common inhabitant of healthy skin, indicating the complex, multifactorial nature of skin infections. Other commonly identified organisms included Corynebacterium, Propionibacterium, Brevibacterium, Streptococcus and Micrococcus spp., with variable results between studies based, in part, on the relative difficulties in isolating some of these organisms.
While less intensively studied, it was also recognized that a commensal fungal microbiota existed, predominantly involving Malassezia spp. in sebaceous regions, along with a limited parasitic population (i.e. Demodex spp.).10,11 Very few studies on commensal viruses have been done, probably due to limitations and difficulties of broad viral screening, yet it is likely that a population of commensal mammalian viruses and bacteriophages exists on the skin.
The bacterial microbiome: culture-independent investigation
With the use of next-generation sequencing, it has become clear that the human skin microbiome is complex, diverse and abundant. For example, a study of the hands of 51 young healthy adults identified an astounding diversity, with hands typically containing >150 different bacterial species.12 Remarkably, the overall bacterial diversity on the hands was reported to match or exceed that found in the mouth, oesophagus and even some small intestinal sites.12 Therefore, while the skin certainly harbours fewer overall micro-organisms per unit of space than the gut, this numerically smaller microbiome can be equally diverse.
Four main bacterial phyla predominate on human skin, Actinobacteria, Firmicutes, Bacteroidetes and Proteobacteria (Tables 1 and 2). Remarkable, interindividual variation is present, probably reflecting differences in local conditions (e.g. temperature, acidity, pH, moisture, hair coverage, nutritional sources), as well as differences in exposure to micro-organisms from the environment, other individuals or other body sites on the same individual. Indeed, the variation between different body sites means that the microbiome from a site on one person’s body is more likely to be similar to the same site on another person than a different ecological niche on the same person (e.g. the microbiome of a person’s forehead is more similar to another person’s forehead than the same person’s hand). Despite this intraperson variation, contralateral sites on the same person tend to be more similar than the corresponding site of a different person (e.g. the left and right hand microbiomes on the same person are more similar to each other than the corresponding hand of another individual).2 The relatively unique nature of an individual’s hand microbiome has even led to investigation of the use of metagenomics for forensic purposes. Comparison of the skin-associated microbiome from objects to the skin microbiome of a person has been used to identify the individual who touched the object, albeit with mixed results.13,14
|Right axilla||Staphylococcus*, 60%||69|
|Back, abdomen,||Streptococcus*, 26%||70|
|chest, limbs and neck||Staphylococcus*, 16%|
|Combined body sites||Corynebacterium†, 23%||71|
Symbols indicate bacterial phyla that correspond to the cited genera, as follows: *Firmicutes and †Actinobacteria.
Marked variation can also occur within an individual over very close distances.2 For example, hair follicles and sebaceous glands are relatively anoxic environments and can therefore support anaerobes that would have little chance to flourish or even survive only millimetres away. Propionibacterium spp. (Actinobacteria phylum) tend to dominate in sebaceous regions, while Staphylococcus spp. (Firmicutes phylum) and Corynebacterium (Actinobacteria phylum) are most common in moist regions.2 Interestingly, dry regions are the most diverse, perhaps reflecting a location less amenable to specific bacterial adaptation and more representative of the multitude of micro-organisms to which the skin may be exposed.
Evaluation of bacterial diversity of different sites raises interesting questions about factors that influence intra-and interindividual variation. In one study, the microbial diversity was higher on the hands of women compared with men, with the microbiome also influenced by the time since hands were last washed and by handedness (i.e. left or right).12 Additionally, while the bacterial diversity may be the same between dominant and nondominant hands, the composition of the microbiome varies.12 The impact of handedness is interesting and probably represents different types of environmental and body site exposure between dominant and nondominant hands. The role of gender is less clear and might relate to the fact that men tend to have a more acidic skin environment,15,16 because increased acidity has been linked to decreased microbial diversity in other ecological environments.17,18 However, this is rather speculative, and other factors, such as sweat and sebum production, use of moisturizers or cosmetics, skin thickness, hormones, environmental contacts and frequency of handwashing, could all play a role.
Variation can also occur based on age. The skin microbiome is established shortly after birth, with vaginally delivered infants acquiring a microbiome similar to their mother’s vagina and infants delivered by Caesarean section acquiring a microbiome most similar to the mother’s skin.19 Not surprisingly, the composition of the microbiome develops gradually over time, with increased relative abundance and evenness of the community developing with age.20 Infants also have different phylum distributions compared with adults, with streptococci and staphylococci accounting for up to 40% of the total microbiome in young infants, and the abundance of the initially low-predominance genera increasing with age.
In addition to the concept of transient and resident flora, the core microbiome needs to be considered. The core microbiome can be defined as ‘the suite of members shared among microbial consortia from similar habitats’.21 In skin microbiome assessment, this represents the micro-organisms that are found in most or all samples from similar sites in different individuals. The core microbiome is presumably rather analogous to the resident microbiome, and it is assumed to be the group of micro-organisms that are critical for proper function of the community. In other ecological niches, the core microbiome tends to consist of a limited number of species or genera that comprise a large proportion of sequences, although by definition core components of the microbiome only need to be common amongst different individuals, not abundant within individuals. There has been limited evaluation of the core skin microbiome, and this population is currently ill defined. A study of forearm and forehead skin determined that 4.5% of genera (Propionibacterium, Staphylococcus and Corynebacterium) and 1.5% of species-level sequences were found in all subjects,22 although as a cloning-based study the overall number of sequences studied was limited, something that may have affected the ability to define members of the core microbiome that were present at low abundances.
Even with the wealth of information generated by broad-range sequence-based studies, methodological limitations need to be considered. For skin, a major variable may be sampling technique. Skin swabs are commonly used because they are easy and noninvasive; however, it must be considered whether they adequately sample the skin microbiome, particularly bacteria that are resident in deep regions such as hair follicles.3 It has been estimated that skin swabs collect approximately 105 bacteria/cm2, while scrapings collect 5 × 105 bacteria/cm2 and skin biopsy specimens 106 bacteria/cm2.3 Relative numbers are less important than differences in the organisms identified, and a study of human skin reported marked similarity in microbial composition data obtained using those three methods, with operational taxon units (groups of similar sequences) found by all three methods accounting for 97% of the overall sequences.3 While some sequences were only found by individual sampling methods, these were low-abundance sequences that may be of limited concern when studying the microbiome. These data need confirmation for other body sites and other species, yet they suggest that easy-to-obtain skin swabs may be adequate.
The fungal microbiome
Minimal investigation of the fungal microbiome has been reported beyond targeted prevalence studies of specific organisms. The Malassezia microbiota has been assessed most thoroughly, with gender, body site, time of year and age identified as affecting prevalence and distribution.23,24 The composition, diversity, abundance and variability of other fungal groups require further investigation.