The human body harbors trillions of microbes colonized mainly in the skin and mucosal surface. The symbiotic co-evolution of commensal microbes with the host is evident from the knowledge that many of these microbial species have given up genes necessary for their survival in other microenvironments, while retaining genes which are beneficial for the host with no/little benefit to self,1 and show diversity between individuals and demographics.2 This diversity is influenced by several factors including mode of birth,3 nutritional patterns, hygiene and life-style (Fig. 1). For example, urbanization and a Westernized diet have led to the loss of ancient commensal microbes. Alterations in microbial diversity are associated with inflammatory conditions including allergies.4 Administration of probiotics has been reported to amend this dysbiotic drift by restoring the gap between the current and ancient symbiont communities.

The Food and Agriculture Organization (FAO)/World Health Organization (WHO) defines probiotics as “live microorganisms which confer a health benefit on the host when administered in adequate amounts.”5 Researchers and regulatory authorities, however, caution against the misuse of the term ‘probiotics.’ They advise that until the beneficial properties of microorganisms are fully characterized, the term ‘probiotic’ should be used with caution. The International Scientific Association for Probiotics and Prebiotics also recently stated that the term probiotics does not apply to dead microorganisms, microbial by-products, metabolites or non-viable microbial products.5 This concept, however, should be revised as effector molecules can be derived from cellular components such as lipopolysaccharides and polysaccharides.67 There is increasing evidence to suggest that every probiotic microbial strain does not have a unique mechanism of action despite shared taxonomic lines.8 Advances in bioinformatics-based research and mechanistic understanding of probiotics' mode of action will perhaps allow some generalization to be made at the taxonomic level with respect to the immunological mechanism of probiotics.8

Studies in the past decade have led to a deeper understanding of the interplay between microbiome and the innate and adaptive immune system to guide immune tolerance and to alleviate allergic disease. We here attempt to review the current evidence and advancements in research for use of probiotics in allergic disorders. Also, as ‘Efficacy and Toxicity’ are interdependent characteristics of therapy, we highlight the need for improved safety regulations.

Primarily, the infant immune system presents type 2 helper T cell (Th2)-dominant phenotypes and gradually increases type 1 helper T cell (Th1) immune response to maintain immune homeostasis. Th2-skewed phenotype leads to higher immunoglobulin E (IgE) levels and mast cell activation, which in turn increases susceptibility to allergic disorders. The microflora plays a significant role in restoring Th1/Th2 immune responses.9 The Th2-dominant phenotype of an infant is reproduced in germ-free (GF) mice, which exhibits higher susceptibility to allergic reactions. Interestingly, colonization with commensal microbes can reverse this phenotype, demonstrating the key role played by microbiota. Commensals also play a role in regulating immune cell recruitment. Neonatal exposure to microbiota prevents accumulation of invariant natural killer cells in the colonic lamina propria and the lungs in a CXCL6 chemokine-dependent manner, thereby ameliorating airway allergic responses.10 Furthermore, loss of tolerance was reported in adults after extensive antibiotic treatment.9

Another perspective of these observations is reflected in ‘hygiene hypothesis.’ It suggests that reduced microbial exposure during infancy due to increased public hygiene is one of the primary causes of heightened susceptibility to allergies.9 This hypothesis is elegantly supported by a recent study in which Amish children raised in conventional farms had lower prevalence of allergic disease than Hutterite children raised in industrialized farms, despite similarities of genetic background.2 Also, house dust extracts from Amish houses inhibited ovalbumin (OVA)-induced allergic inflammation in mice in a MyD88 and TRIF-dependent mechanism.2 Of note, the bacterial composition of Amish house dust was dominated by Bartonellaceae, which had higher endotoxin levels than Hutterite house dust. Long-term exposure to low-dose endotoxin or ‘farm dust’ demonstrated protective effects in house dust mite (HDM)-induced asthma in mice.11 Protection was mediated via activation of A20 (TNFAIP3) in airway epithelial cells modifying the epithelial and dendritic cells (DCs) interactions.11 Similarly, these studies establish the role of microbiota in influencing allergic immune reactions. In the subsequent part of the review, we elaborate on probiotics-based intervention of allergic diseases and propose changes required for the advancement of the probiotics field.

Inflammation is a fundamental defense response of body's immune system against foreign antigens; however, hypersensitivity is a host protective immune response on repeated exposure to specific antigens, but potentially deleterious to the host. Based on physiological mechanisms involved and time of response, Gell and Coombs have classified hypersensitivity into 4 types, types I to IV. However, the clinical manifestations of allergic reactions demonstrate overlap or simultaneous occurrence of humoral and cell-mediated immune responses making the differential diagnosis of 4 types difficult. Johansson et al.12 proposed a more clinically applicable classification based on involvement of IgE, excluding non-allergic non-immunological reactions. A general pathogenesis of an allergic response includes: sensitization to allergens (HDM, pollen, food, etc.), transient inflammatory response, increase in allergen specific IgE, and migration of effector T cells, mast cells and eosinophils to the site of exposure. Hyper-responsiveness is orchestrated by Th2 cells, with cytokine profiles dominated by interleukin (IL)-4, IL-5, IL-9, IL-13, and IL-31. Eosinophils play a major role in respiratory allergies by degranulation at the site of exposure causing influx of immune cells and tissue damage. Skin and food allergies can also be referred to as the second wave of allergic epidemics.

As described above, although health benefits of taking probiotics have been reported in allergic disorders, it is still early to draw any conclusions. Prominent regulatory organizations like the American Academy of Pediatrics, National Institute of Allergy and Infectious Diseases, European Academy of Allergy and Clinical Immunology, European Society for Pediatric Gastroenterology, Hepatology and Nutrition and FAO of the United Nations/WHO, do not support the use of probiotics for primary prevention of allergic diseases.51 In 2015, the World Allergy Organization (WAO) applied Grading of Recommendations, Assessment, Development and Evaluation approach to develop evidence-based recommendations for using probiotics in the prevention of allergic diseases.52 Findings were reported in a systemic review of 29 RCTs by Cuello-Garcia et al.43 The panel observed that there is insufficient evidence to support recommendation to use probiotics in primary prevention of allergic diseases. However, they suggested the use of probiotics in pregnant/lactating women and infants with a family history of allergic disease. Although strain-specific activity of different bacterial species is recognized, no recommendations were made by the WAO regarding strain or dose of probiotics in light of insufficient evidence.52

Probiotics are generally considered safe for consumption. However, increasing evidence raises concern over generalizing the safety of probiotics. In 2002, a joint report released by WHO and FAO of the United Nations indicated that probiotics may be responsible for 4 types of side effects: systemic infections, deleterious metabolic activities, excessive immune stimulation in susceptible individuals and gene transfer.53 The ‘Probiotics in Pancreatitis Trial (PROPATRIA),’ a prophylactic clinical trial for acute pancreatitis in human patients, created a furor as large number of deaths were reported in the probiotic-treated group compared to control, 16% versus 6%.54 It was also reported that treatment of patients with a multi-strain probiotic led to bowel ischemia development.54 An enquiry into the trial identified inadequate knowledge on the safety of probiotics used. This investigation grabbed the attention of the scientific community towards looking into adverse effects of probiotics. Lactobacillus and Bifidobacterium are generally considered safe, and therefore dominate the commercial formulas of probiotics. A recent investigation, however, identified Bifidobacterium adoloscentis as a potent inducer of IL-17, which plays pathogenic roles in inflammatory bowel disease (IBD) and rheumatoid arthritis.55 Although Bifidobacterium species have long been associated with Treg induction,22 including Bifidobacterium adolescentis,55 the current findings indicate that a harmless probiotic strain can be counterproductive in disease conditions. Tan et al.55 further tested 6 probiotic formulas containing Bifidobacterium species, 4 of which similarly induced Th17 cells in mice, although their effects in an autoimmune set-up were not confirmed. These observations warrant an investigation into the clinical safety of probiotics before commercialization. Some groups further argue that probiotics with medical claims must receive Food and Drug Administration (FDA) approval like any other drugs, including filing of Investigational New Drug Application (IND) and clinical trials, phases I–III. Other groups support the necessity of regulations, but disagree on receiving FDA approval as this will incur huge expenses and cause major delays in the overall research.5 We believe the safety of probiotics must not be overlooked; the IND application process can be modified for probiotics products to expedite research without adding much to the cost.

Advances in molecular and microbiological techniques have enabled us to perform in-depth research into microbial diversity found in the human commensal pool. To-date, however, most of the probiotic research has focused on the bacteria itself rather than the host-bacteria interaction. Moreover, available data omits the importance of mycobiome and virome. This has arisen from difficulty in processing fungal DNA, underdeveloped fungal genome databases, unavailability of viral 16S rRNA sequencing and inefficient bioinformatics tools. In our observations, commensal yeast, recognized mostly as an opportunistic pathogen, can act as a potent immunoregulator in inflammation and autoimmune diseases (unpublished data). The interaction and harmony between microbiome, mycobiome and virome can be important in maintaining homeostasis and offers a window for further investigation.

As more academic and clinical research groups focus on commensals, the number of probiotic species with health claims is increasing. However, there is no international consensus for the screening of these microbial species. Current screening methods focus on the cytokine induction ability of microbes using cell lines or ex vivo isolated peripheral immune cells, even though they do not represent the phenotype of intestinal cells. There is a need to devise high-throughput screening methods to ensure the specificity and efficacy of selected probiotics. Most commercially available probiotics are mixtures of bacterial strains with defined colony forming units (CFUs); therefore, it is also important that consumers be informed of the product shelf life of individual strains to ensure viability of bacteria in the administered dosage. A recent report also suggested the manufacturing process of probiotic mixtures can affect their activity. VSL#3, a probiotic mixture made in the USA, was found to be beneficial in HIV-infected subjects and also IBD by reducing the 1,3 dihydroxyacetone (DHA) levels in feces.56 On the contrary, a similar mix made in Italy enhanced DHA levels leading to adverse effects.56 These findings suggest the same bacteria can have different functions, depending on their culture conditions. Thus, using functional markers rather than CFU counts might be a better readout of probiotic quality control. Moreover, as the composition of commensals varies among populations, efficacy of probiotics may also vary among populations. Hence, clinical trials should be expanded to include different geographical locations. With this in mind, it is advisable not to perform meta-analyses on pooled data when different strains of bacteria were used, since the effects can vary dramatically among strains.55

The majority of disease-related health claims for probiotics have been rejected due to lack of supportive data. Then, how can we establish probiotics as a modulator of specific diseases? Analysis of disease-associated markers could be a simple approach. For example, Roessler et al.57 explored different effects of selected probiotics in healthy and AD patients, and found probiotic treatment increased CD57⁺ NK cells in healthy subjects, but not in AD patients. Kim et al.58 reported that characterization of clinical phenotype of patients and selection of patients based on the immunological parameters (i.e., high IgE level) could enhance the possibility of probiotic efficacy. However, these approaches do not cover many important issues such as the molecular and cellular identity of the probiotic strain, full genome sequences and mechanisms of action in homeostatic/disease conditions. Therefore, application of probiotic strains in patients should not be allowed unless the safety and effector molecules of the probiotics are well defined. Moreover, the term “Probiotic” should not be generically used as some exacerbate disease under different conditions.

The current understanding of probiotics suggests their beneficial effects to prevent AD. However, evidence is still lacking to support the general efficacy of probiotics in the modulation of other allergic conditions. Valid conclusions cannot be drawn based on pooling of data from different probiotic strains or combinatory administrations. Thus, there is a need to develop international standards for study designs on probiotics to ensure uniformity across clinical trials. Further, there is a glaring deficiency in quality standards of probiotics which shall be addressed by defining genetic identity of strains and uniform production guidelines. Moreover, future studies are warranted to refine the effector molecules of probiotics and to identify their mode of action in healthy and diseased conditions. We also want to emphasize the need of functional markers for screening probiotics, which could serve as a standard for quality control toward each batch of manufactured products. Annotating genomic information involved in coding the functional molecules is also necessary as bacteria could lose their genetic information during large-scale cultures. Once the safety and efficacy of the effector molecule(s) are assured, application of probiotic bacteria for treating allergic disorders might be possible. In addition, once they show beneficial effects in clinical trials, we could call them “Probiotics.” Furthermore, if we can purify the effector molecules from the probiotic bacteria in large amounts, we may use this to treat allergy symptoms in the future (Fig. 3). Although, current definition excludes these products from the category ‘Probiotic,’5 we propose that microbial derived molecules with health benefits could be included in a separate category within probiotics, termed “Pharmabiotics.”