Introduction

Acne vulgaris is the most common dermatological disorder worldwide. It affects around 50 million people each year in the USA, with an estimated annual cost of $2.5 billion [1]. The worldwide prevalence of acne is estimated to be around 9% [2], accounting for 0.3% of the global disease burden [3]. Although acne affects people of all ages, 85% of all affected individuals are 12–24 years old [4,5,6,7]. Severe acne also carries a high social and psychological impact, affecting emotions, self-esteem, and increasing the risk of depression and suicide [7].

The role of Propionibacterium acnes in the pathophysiology of acne is still under debate. P. acnes is the predominant commensal microorganism of the human skin microbiome. A delicate balance within the skin microbiota is essential for the barrier function of the skin and prevention of pathogen colonization [8].

The emergence of antibiotic resistance has become a public health problem worldwide [9]. The long-term use of topical and systemic antibiotics has led to high rates of antibiotic-resistant P. acnes strains [10]. As research on new antibiotic agents is decreasing due to cost and difficulty, the development of new, natural, and non-conventional alternatives—such as antimicrobial peptides, natural engineered antibodies, and bacteriophages—is becoming critical. Bacteriophage therapy seems to be a promising alternative. Its advantages are host specificity and simplicity of isolation and production. Although both in vitro and in vivo studies have shown the potential of targeted bacteriophage therapy in skin infections, research is lacking on bacteriophage therapy targeting P. acnes-associated infections. It has recently been proposed that the species P. acnes be reclassified to Cutibacterium acnes and other genera [11]. Here, we use the old nomenclature (P. acnes) throughout because it is still used by most of the evidence presented in this review. In this review, we summarize the most recent evidence on P. acnes bacteriophages and its potential role in the treatment of acne. This article is based on previously conducted studies and does not contain any studies with human participants or animals performed by any of the authors.

Acne Vulgaris Pathophysiology

Acne is a multifactorial disease. Increased sebum production by androgen stimulation, abnormal hyperkeratinization of the pilosebaceous duct, and subsequent bacterial colonization and inflammation all contribute to the disease [12, 13]. It is proposed that P. acnes colonization plays a pivotal role in the pathogenesis of acne since antimicrobial therapy has been effective in treating acne for many years. However, its contribution to acne development is controversial [14].

Propionibacterium acnes is a Gram-positive, anaerobic/microaerophilic, fat-splitting, rod-shaped bacterium found on the skin; it represents nearly 90% of the skin microbiome of healthy adults [1, 14, 15]. The concentration of P. acnes depends on the abundance of sebaceous follicles and the age of the individual [13, 16, 17]. Accordingly, its concentration is higher on sebaceous areas such as the face, scalp, and back [13, 18], and various studies have reported an association between P. acnes levels and sebum production [16]. There is a marked increase in P. acnes colonization during puberty [15], which correlates with the time when sebaceous glands mature [13]. P. acnes may disrupt keratinocyte differentiation in the follicle, thereby contributing to the formation of comedones and inflammatory acne lesions by triggering a host inflammatory response [12, 19]. P. acnes produces enzymes that degrade skin components as well as chemotactic factors that stimulate keratinocytes and inflammatory cells to release pro-inflammatory cytokines (e.g., interleukin [IL]-8, IL-12, IL-1α, IL1-β, tumor necrosis factor alpha) and reactive oxygen species [1, 13, 20,21,22,23]. Although P. acnes is a commensal organism in humans, not all healthy adolescents or adults develop acne, indicating that differences in the pathogenicity of P. acnes strains must exist [1]. It has been proposed that certain strains play a pathogenic role and others act as bystanders [24,25,26,27]. Studies have also shown that specific genes in the P. acnes genome contribute to bacterium virulence and hence to acne pathophysiology [14, 26, 28].

Emerging Antibiotic Resistance

Over 2 million Americans become infected every year with antibiotic-resistant bacteria, resulting in about 23,000 deaths [29, 30]. The post-antibiotic era is approaching as antibiotic effectiveness steadily declines, and multiple common infections become resistant to treatment [14, 31]. The excessive use of antibiotics in agriculture and humans, the evolutionary pressure inherent to antibiotics [31], and the lack of research on new antibiotic agents are some of the reasons behind antibiotic resistance [14, 29, 31, 32].

Antibiotic resistance is among the main causes of treatment failure in acne vulgaris [33]. The mechanism of antibiotic resistance to P. acnes is explained by the remarkable genetic plasticity of bacteria [34]. Two major genetic strategies permit antimicrobial resistance: (1) gene mutation, and (2) foreign DNA coding acquisition through horizontal gene transfer [34]. Gene mutation is the predominant mechanism leading to P. acnes antibiotic resistance [12, 35].

Acne therapy warrants long-term treatment with topical and systemic antibiotics, which contributes to resistant P. acnes strains. Globally, P. acnes resistance to antimicrobials has increased almost 40% between the 1980s and 2000s worldwide [10]. Erythromycin/clindamycin-resistant P. acnes seems to be the most common pattern of resistance based on reports from the USA, Europe, and Asia [36,37,38]. However, resistance rates vary by region. Europe, Singapore, and Hong-Kong have high prevalence rates of erythromycin/clindamycin-resistant P. acnes (45–91%) and tetracycline-resistant P. acnes (2–26%) [36, 37, 39,40,41,42], but countries that practice conservative use of antibiotics, such as Japan and Korea, report much lower resistance rates (2–4%) [36, 42, 43].

The long-term use of antibiotics can promote the formation of an antibiotic-resistant biofilm that protects the bacterium against host defenses and can alter the natural microbiota of the skin [12, 44]. Studies have shown colonization by antibiotic-resistant coagulase-negative Staphylococci and Streptococcus pyogenes in acne patients who have used both topical and oral antibiotic therapy [45, 46]. Thereby, it is recommended to limit monotherapy with topical antibiotics and instead to combine them with other topical agents such as retinoids or benzoyl peroxide to decrease the risk of resistance [47].

Human Skin Microbiome

The human skin microbiome is home to a variety of microorganisms, including bacteria, fungi, viruses, and arthropods. A delicate balance between the microorganisms is essential for local immunity and barrier function of the skin [8]. Imbalances in this system have been linked to dermatologic diseases, such as acne, atopic dermatitis, psoriasis, and rosacea [48].

The dominant bacterial species found on adult skin are Propionibacterium, Corynebacterium, and Staphylococcus [48]. The skin can be divided into dry, moist, or sebaceous microenvironments, and each of these microclimates host varying proportions of these common bacterial flora [48].

Age-related shifts in bacterial communities could explain why certain skin diseases are prevalent at different stages of life. For example, the microbiome in children, who are more susceptible to atopic dermatitis, is composed of mostly Streptococcaceae, Bacteroides, and Proteobacteria [49]. Pubertal spikes in androgens lead to a more susceptible environment for the development of acne and a shift in microbiome concentration to greater levels of Propionibacterium and Corynebacterium.

Bacteriophages

Delving deeper into the bacteria of the microbiome brings us to bacteriophages. Bacteriophages, or viruses that infect bacteria, can be found throughout the biosphere, are essential members of the human microbiome, and may play an important regulatory role in human skin health and disease [8, 50,51,52]. Little is known about bacteriophage interaction with skin microbiota. Bacteriophages are obligatory intracellular parasites; thereby, their distribution depends on their host organisms [50]. There are over 6000 well-known bacteriophages [53], and these bacteriophages are estimated to be at least tenfold more common than bacteria [14]. There is wide diversity in the structure of these phages (e.g., tailed, polyhedral, pleomorphic, filamentous), and they are usually classified based on their genetic content [53].

Four distinct life cycle phases have been described: lytic (virulent phages), lysogenic, pseudo-lysogenic (temperate phages), and chronic infection [53] (Table 1). The first step in every phage cycle is the binding of the phage to bacterial surface receptors, after which the phage injects its genetic material (DNA or RNA) into the cells [54]. Phages undergoing the lytic phase, also called virulent phages, are the most abundant type and the most widely used in bacteriophage therapy due to their natural ability to kill bacteria directly through cell lysis.

Table 1 The cell cycle of bacteriophages
Full size table

Phages (temperate or dormant phages) that enter the lysogenic phase have the ability to induce transduction [53], a vital process in bacterial pathogenesis and adaptation [29, 55]. Transduction can confer advantages to bacteria by transferring pathogenicity or antibiotic resistance genes [14, 53, 56, 57]. Due to this risk, only those bacteriophages with lytic activity should be considered in phage therapy [53]. However, transduction can be used to our benefit by genetically engineering phages to transfer genes to reverse antibiotic resistance or to increase bacterial susceptibility to antibiotics [29]. Edgar et al. restored antibiotic sensitivity to streptomycin and nalidixic acid in resistant Escherichia coli in an in vitro study using genetically engineered bacteriophages [58]. In another study, Lue and Collins used genetically engineered phages to transfer genes that target DNA repair mechanisms, resulting in an increase in E. coli susceptibility to quinolones in vitro and in vivo [59]. These studies show that engineered bacteriophages can enhance the killing of antibiotic-resistant bacteria and biofilm and reduce the emergence of antibiotic resistance in bacteria.

Propionibacterium acnes Bacteriophages

Propionibacterium acnes bacteriophages were first identified by Brzin in 1964 [60]. In 1968, Zierdtc et al. isolated phage 174 from Corynebacterium acnes strains and used it to classify the Corynebacterium family, ultimately finding that 88% of all C. acnes strains were sensitive to this bacteriophage [54, 61]. A later study reported that 18% of P. acnes phages carried bacteriophages [62], and P. acnes bacteriophages were subsequently used to classify Corynebacterium and Propionibacterium [63, 64].

P. acnes bacteriophages are relatively more abundant in lipid-rich areas of the skin, correlating with the distribution of P. acnes in the skin [65,66,67]. These bacteriophages are the dominant phages in the pilosebaceous unit. Fitz-Gibbon et al. reported a 1:120 bacteriophages:P. acnes ratio in pilosebaceous units in healthy skin samples [26, 68, 69]. Most P. acnes bacteriophages possess a siphoviral morphology (i.e., isometric head and long flexible tail) [8, 15, 70, 71] and have a pseudolysogenic life cycle (Table 1) [8, 15, 72, 73]. P. bacteriophages displaying a lytic life cycle have also been characterized [70, 74].

Interestingly, despite the isolation of P. acnes bacteriophages over a varied temporal and geographical range, their genome is preserved with very limited genetic diversity [8, 15, 71, 72]. Marinelli et al. investigated the diversity of bacteriophages that infect P. acnes and isolated 11 P. acnes bacteriophages which lacked the genetic diversity seen in other phage populations [15]. A recent study by Liu et al. sequenced 48 P. acnes bacteriophages from human skin follicles and found a sequence identity of between 85 and 100% between strains, suggesting that the P. acnes bacteriophage population in the skin microbiota is dominated by one strain [8]. The authors tested the P. acnes–bacteriophage interaction and found that the 74 P. acnes strains were susceptible to the 15 tested P. acnes bacteriophages. They suggested multiple reasons for the lack of phage diversity, including a bottle-neck hypothesis leaving one dominant genotype, or the evolutionary constrains imposed on phages and bacteria to maintain a single phage, thus limiting the spreading of phage resistance [8]. Another possible explanation for the limited diversity of P. acnes bacteriophages could be due to the niche in which they live, as P. acnes makes up 90% of the microbiota of the pilosebaceous unit, thereby limiting horizontal gene transfer and increased diversity between phages in the pilosebaceous unit [14].

Liu et al. also found that some individuals shared the same bacteriophage strains in the skin microbiota, suggesting the existence of a pool of common bacteriophages among human populations [8]. They further discovered identical bacteriophages strains between closely related individuals (siblings), which makes human to human virus transmission a possibility. As seen in other studies [15], the resistance of certain P. acnes strains to bacteriophages was an issue [8]. Two possible resistance mechanisms are described by these authors, namely, restriction modification and clustered regularly interspaced short palindromic repeat (CRISPR), both of which target viral DNA integration into the host genome [8]. The findings of this study led the authors to conclude that the ability of P. acnes bacteriophages to lyse only susceptible strains may alter the bacteria population, as different strains will grow at different rates, thereby modulating the composition and dynamics of the skin microbiota.

The apparent lack of genetic diversity of P. acnes bacteriophages and their broad host range make them ideal candidates for phage therapy in acne [14]. Moreover, lytic bacteriophages engineered to target P. acnes strains in the specific microbiome of individuals will increase the success rate of acne treatments.

Propionibacterium acnes Bacteriophage Therapy

The potential role of bacteriophage therapy in acne vulgaris has recently attracted the interest of researchers and clinicians. Phages active against P. acnes have been isolated from the skin, oral cavity, and gastrointestinal tract [14]. It is important to note that only phages with proven lytic activity should be used in phage therapy because lysogenic or temperate phages carry the risk for transduction of antibiotic resistance or pathogenicity genes and may lead to delayed cell lysis [14, 53].

Bacteriophage therapy has been used in humans for several types of infections with good results [75,76,77,78]. However, no trials on P. acnes bacteriophage therapy have been conducted in humans. Brown et al. isolated ten bacteriophages capable of lysing P. acnes from human skin microbiota and tested their therapeutic potential [72]. These authors created a suspension for each bacteriophage at a final concentration of 2.5 × 108 PFU/g using an aqueous cetomacrogol cream that showed that these bacteriophage formulations effectively lysed P. acnes cells in agar lawn culture plates and remained active in the cream for up to 90 days when stored at 4 °C in light-protected bottles. The bacteriophage was specific to P. acnes strains and did not lyse other bacteria of the Propionibacterium family. Cells that regrew from the areas within the P. acnes plaques showed phage resistance [72]. Although some authors have suggested that a cocktail of phages could be used to decrease the risk of phage resistance [8, 15], P. acnes bacteriophage variability is relatively low, which may limit this approach. This important limitation needs to be explored in further studies. However, the results using the cream formulation of Brown et al. [72] suggest that P. acnes bacteriophage therapy is a simple and realistic therapeutic option for the treatment of acne. Another in vitro study showed effective eradication of P. acnes strains when P. acnes bacteriophages were isolated from human skin microbiota and applied in drops onto agar plates [79]. In this study, P. acnes bacteriophages were unable to kill other bacteria, such as Staphylococcus aureus, S. epidermidis, and Corynebacterium xerosis, confirming the specificity of these bacteriophages [79]. Formulations such as oil–base cream, water–oil nanoemulsion, biodegradable polyester matrix, antiseptic gel, and paraffin-oil-based lotion, have proven to be effective strategies to deliver bacteriophages [80].

P. acnes bacteriophage genomes encode endolysins involved in bacteria cell-wall degradation (muramidases, amidases, endopeptidases, glucosaminidases, and transglycosylases) [15, 70, 81]. These endolysins are implicated in the release of progeny following phage assembly by targeting peptidoglycan in the bacterial wall [70]. Marinelli et al. suggested that endolysins are a potential therapeutic option in acne therapy [15]. Phage endolysins are highly conserved in different P. acnes bacteriophage strains (95% at the amino acid level) [15], which implies that endolysins from any P. acnes bacteriophage could be active against most P. acnes strains. Phage endolysins have been used as antimicrobials both in vitro and in vivo with promising results [82]. Furthermore, no resistance to phage endolysins has been reported [82]. It has been shown that even bacteria that become phage resistant may remain endolysin sensitive [14]. This introduces yet another way to treat acne through the genetic engineering of enzymes to target bacteria cell walls.

These possibilities carry important therapeutic implications in the management of acne. Antibiotics, in combination with bacteriophage cocktails, could be used to decrease antibiotic resistance and to treat antibiotic-resistant P. acnes. However, further research is needed to evaluate P. acnes bacteriophage therapy in human subjects, both as monotherapy and in combination with conventional therapies.

Advantages of Phage Therapy

Bacteriophages have a low environmental impact compared to chemical antibiotics due to their natural origin [53, 83]. They target both Gram-positive and Gram-negative bacteria [84,85,86,87,88,89,90,91,92], and many in vitro and in vivo models have shown that bacteriophages are effective against multidrug-resistant bacteria [85,86,87,88]. As antibiotic resistance grows, phages retain the ability to kill antibiotic-resistant bacteria due to their differing mechanisms of action [53]. Bacteriophages are specific to their bacterial hosts (species), and only replicate locally, limiting the pressure on normal non-targeted flora of the skin and other organs [75, 93, 94]. Bacteriophages have also been shown to distribute in good concentrations all over the body, including the central nervous system [53, 91, 95] (Table 2).

Table 2 Advantages and disadvantages of phage therapy
Full size table

Another potential benefit of bacteriophage therapy is the ability to decrease biofilm formation [83, 96,97,98,99,100,101]. Many in vitro and in vivo studies have proven that the combination therapy of antibiotics and lytic bacteriophages displays synergism by improving the efficacy of bacteria and biofilm eradication and preventing the emergence of resistant bacteria [92, 102,103,104,105,106,107,108,109,110,111]. Antibiotics could be conjugated with bacteriophages to deliver antibiotics to specific bacteria and at higher concentrations [14]. Furthermore, engineered bacteriophages could be used to improve efficacy through the transfer of susceptibility or sensitizing genes by means of genetic engineering.

The identification of bacteria and bacteriophage isolation for therapeutic purposes is a rapid and affordable process compared to the development of new antibiotics [93]. Moreover, the cost of bacteriophage therapy seems to be lower than that of traditional antibiotic therapy [112]; however, more studies are needed to establish the real short- or long-term costs of bacteriophage therapy.

Finally, bacteriophages are safe and well-tolerated, and no significant adverse events have been reported [75, 77, 78, 94, 113, 114].

Limitations of Bacteriophage Therapy

Although the prospect of using bacteriophage therapy to treat acne in a world with increasing antibiotic resistance is promising, this novel therapeutic endeavor comes with limitations. The first of these is our evolving understanding of phages and their life cycles. The newest data suggests that phages exist on a continuum between lytic and lysogenic life cycles [115]. This creates challenges when using phages as therapeutic vehicles since conventional phage therapy requires phages to undergo lytic cycles and rapidly kill their hosts. Most P. acnes phages characterized thus far, however, display pseudolysogeny (Table 1).

In addition, CRISPR protects bacteria from viral DNA integration, and P. acnes may become resistant to phage therapy through this mechanism [54]. This issue of resistance becomes even more likely with the knowledge that bacteriophages targeting P. acnes are highly homogeneous [26]. Therefore, the acquisition of resistance to one phage may confer P. acnes with resistance to many of its bacteriophages [14]. However, the risk is low compared to antibiotics, partially because bacteriophages can mutate and bypass bacteriophage resistance mechanisms [93]. Resistance can also be prevented by using multiple bacteriophages (cocktails) or synergistic combinations of bacteriophages + antibiotics [93].

In addition to these limitations, the more practical aspects of establishing optimal therapeutic doses, treatment frequency, and duration have not been established [54]. A more long-term risk with this therapy includes the unknown consequences to the cutaneous microbiome if P. acnes, a vital member to this community, is temporarily eradicated through acne treatment. We do not fully understand the repercussions of altering the natural micro-ecosystem of the skin.

Future therapeutic option

As the antibiotic resistance era approaches, research on new alternative antimicrobial agents is becoming critical. Bacteriophages, ubiquitous microorganisms of the human skin microbiome, contain distinct advantages that mark them as promising alternatives to conventional antimicrobial therapy. Among these advantages are host specificity, limited cross resistance, ease of isolation, low cost, and a favorable safety profile compared to antibiotics.

The potential clinical application of bacteriophage therapy for acne vulgaris is promising. Limitations to phage therapy, such as the risk of transduction of pathogenicity genes and the low P. acnes bacteriophage variability, can be overcome by a more thorough understanding of the bacteria–bacteriophage interaction in the human skin microbiome. Phage resistance is another limitation that must be considered and warrants further study.

As the field develops, more data is needed before phage therapy in human subjects is introduced. Developing targeted phage therapy, engineered bacteriophages, and enzyme-based therapies, either alone or as an adjuvant to antibiotics, may lead to decreasing rates of P. acnes resistance to antibiotics and the restoration of antibiotic susceptibility to P. acnes.