Gut Microbiota: The Missing Link Between Helicobacter pylori Infection and Metabolic Disorders?

Helicobacter pylori (H. pylori) is a gram-negative bacterium that infects approximately 4.4 billion individuals worldwide. Although the majority of infected individuals remain asymptomatic, this bacterium colonizes the gastric mucosa causing the development of various clinical conditions as peptic ulcers, chronic gastritis and gastric adenocarcinomas and mucosa-associated lymphoid tissue lymphomas, but complications are not limited to gastric ones. Extradigestive pathologies, including metabolic disturbances such as diabetes, obesity and nonalcoholic fatty liver disease, have also been associated with H. pylori infection. However, the underlying mechanisms connecting H. pylori with extragastric metabolic diseases needs to be clarified. Notably, the latest studies on the topic have confirmed that H. pylori infection modulates gut microbiota in humans. Damage in the gut bacterial community (dysbiosis) has been widely related to metabolic dysregulation by affecting adiposity, host energy balance, carbohydrate metabolism, and hormonal modulation, among others. Taking into account that Type 2 diabetic patients are more prone to be H. pylori positive, gut microbiota emerges as putative key factor responsible for this interaction. In this regard, the therapy of choice for H. pylori eradication, based on proton pump inhibitor combined with two or more antibiotics, also alters gut microbiota composition, but consequences on metabolic health of the patients has been scarcely explored. Recent studies from our group showed that, despite decreasing gut bacterial diversity, conventional H. pylori eradication therapy is related to positive changes in glucose and lipid profiles. The mechanistic insights explaining these effects should also be addressed in future research. This review will deal with the role of gut microbiota as the linking factor between H. pylori infection and metabolic diseases, and discussed the impact that gut bacterial modulation by H. pylori eradication treatment can also have in host’s metabolism. For this purpose, new evidence from the latest human studies published in more recent years will be analyzed.


INTRODUCTION
Helicobacter pylori (H. pylori) is a microaerophilic spiral-shaped Gram-negative bacterium that can colonizes the stomach. This bacterium is able to generate an alkaline microenvironment that allows its survival in the acid gastric environment (1). The prevalence of H. pylori infection is more than 50% worldwide, but ranges from 19 to 88% depending on geographical area localization, with a higher prevalence in developing than in developed countries (2), which has been related to socioeconomic status and hygiene levels, including level of urbanization, sanitation, and access to clean water (2). However, other factors such as racial group could also be related to H. pylori prevalence (2,3). The high prevalence of this infectious disease is a matter of concern due to the high number of H. pylori-associated clinical conditions.
H. pylori infection disturbs gastric homeostasis and, although most of the infected individuals remain asymptomatic, serious gastrointestinal complications are attributed to this bacterium such as peptic ulcer or chronic gastritis and promotes other lifethreatening complications such as gastric cancer and mucosaassociated lymphoid tissue lymphomas (4). H. pylori-related diseases are not only restricted to the gastrointestinal tract, but H. pylori infection has also been associated with a number of extra-gastroduodenal disorders. In fact, there are solid evidence on its relationship with hematological disorders such as idiopathic thrombocytopenic purpura, iron deficiency anemia and vitamin B12 deficiency (5), but there is also a growing body of evidence that supports a link with neurological (e.g. Parkinson or Alzheimer disease, age-related cognitive decline), cardiovascular, liver, metabolic (e.g. diabetes, obesity, metabolic syndrome) and autoimmune and inflammatory (e.g. autoimmune gastritis, immune thrombocytopenia purpura, autoimmune thyroid diseases, inflammatory bowel diseases) disorders (6).
The precise mechanisms underlying the connection between gastric H. pylori infection and extra-gastroduodenal diseases remain unclear. The most accepted hypothesis to date is the generation of an inflammatory milieu due to the H. pylori insults to gastric mucosa and the consequent activation of innate and adaptive responses (1,4). It has been hypothesized that this local inflammation at the stomach can spread systematically by the release of proinflammatory cytokines. This would favor the establishment of a low-grade and chronic inflammation, that is a common feature of H. pylori-associated extra-gastroduodenal disorders including cardiometabolic diseases such as diabetes, atherosclerosis or dyslipidemia (7,8). However, this hypothesis has not been formally confirmed as yet, and knowledge about how inflammation actually links H. pylori with metabolic disorders is still insufficient.
Prophylactic recommendations to prevent H. pylori infection such as good household hygienic practices and the use of safe supplies of clean water are not always feasible, and while awaiting for an efficient H. pylori vaccine, the antibiotic therapy is the strategy of choice to fight against H. pylori infection at present (25,26). Therapies for H. pylori eradication are based on the use of different antibiotic combinations (27)(28)(29) which are well known to have profound effects on the composition and diversity of the gut microbiota (30) and might also be related to adiposity and insulin resistance (31)(32)(33)(34). Although an increasing number of studies have emerged analyzing the effects of the various antibiotic eradication therapies for H. pylori eradication on the composition of the gut microbiota (10,11,16,17,24,(35)(36)(37)(38)(39)(40)(41)(42)(43)(44), few of them have related these changes with patients' metabolic traits (10,(22)(23)(24). Within this context, our group has recently focused on analyzing how gut microbial changes promoted by H. pylori and its treatment can relate to metabolic traits.
In this review, we discuss the latest evidence from human studies on the influence of H. pylori infection and different eradication therapies on the composition of the gut microbiota, with particular focus on the relationship between H. pyloriinduced gut microbiota modifications and host's metabolic health.

H. PYLORI INFECTION AND METABOLIC DISEASES
Among the H. pylori infection-derived extra-gastric manifestations, the link with energy management, weight gain and metabolic homeostasis is still up for debate (6).
On the one hand, alterations in the gastrointestinal microenvironment due to H. pylori infection have been suggested to impair nutrient absorption and causes micronutrient deficiency. Malabsorption of iron and vitamins such as vitamin B12, vitamin A, vitamin C, vitamin E or folic acid have been related to H. pylori infection. Concordantly, associations between micronutrient deficiency complications and H. pylori infection have been found. For instance, the incidence of iron and B12 deficiency anemias is higher in H. pylori positive patients than in non-infected subjects (45,46). Other disorders related to micronutrient deficiencies have been suggested to be associated with H. pylori infection, although further evidence would be necessary to confirm this hypothesis (45,47,48).
Moreover, lower levels of the orexigenic gastric hormone ghrelin (49), and higher levels of the anorexigenic adipokine leptin (50) have been reported in H. pylori infected patients. All together, these data led to propose that H. pylori infection can cause growth retardation and malnutrition in children. However, studies found that growth delay was also dependent on socioeconomic factors (8). Findings on weight management in adults during H. pylori infection are controversial with studies having reported positive (51)(52)(53), negative (54,55) or non correlation (56,57) between H. pylori infection and increased body mass index (BMI).
H. pylori eradication has been associated with weight gain likely due to the improvement of postprandial dyspeptic symptoms. In addition, changes in ghrelin and leptin levels after H. pylori treatment could be also acting on weight increase (58)(59)(60). An intervention study carried out in H. pylori positive children with previous growth retardation and low ghrelin levels, showed restoration of normal ghrelin levels and increased weight gain upon H. pylori eradication (61). By contrast, other studies reported significant weight loss, decrease in fat mass percentage and increase in fat-free mass after H. pylori eradication treatment (62,63) or no effect on body weight (64) in adults. These contradictory results can be due to the different populations analyzed including different age groups. The influence of H. pylori treatment on body weight seems to differ in adults and children which could be due to age-related phenomena on body growth. In addition, other factors, such as gut microbiota, could be determining these discrepant results as discussed below.
On the other hand, imbalance of the aforementioned and other hormones such as glucagon-like peptide 1 (GLP-1), and the generation of a proinflammatory milieu might negatively affect metabolic homeostasis and weight in the adulthood (8). In fact, low-grade inflammation is a common feature of obesity, diabetes, insulin resistance, dyslipidemia and cardiovascular diseases (65).
Patients suffering of H. pylori infection are prone to display an unfavorable pro-atherogenic lipid profile featured by high triglycerides, total cholesterol and LDL-C and decreased HDL-C levels (22,(66)(67)(68)(69). Zhao et al., found that, in addition to the dyslipidemic profile, bilirubin levels were also diminished in H. pylori infected patients (70). Bilirubin exhibits powerful antioxidant properties and has been inversely related to the risk of cardiovascular diseases and metabolic syndrome (71,72). However, some studies did not find correlations between H. pylori infection and some of these lipid variables (73,74). A recent meta-analysis of 27 previous studies concluded that most of the evidence tends towards an unfavorable lipid profile in H. pylori infected patients, although this is still casting doubt on the precise relationship with triglyceride levels (75). The causes of the altered lipid profile have not been elucidated, but impaired intestinal absorption and altered bile acid dynamics might be contributing to this condition (76). Notably, as discussed below, gut microbiota modulates these processes and could be acting as mediators in the establishment of the dyslipidemic profile in H. pylori patients. Supporting this hypothesis, antibiotic treatment for H. pylori eradication resulted in total and LDL cholesterol reduction and increase in HDL-C levels (77). Other studies described an elevation in total cholesterol levels at the expense of increasing HDL-C after H. pylori eradication (22,78). However, the lipid improvement after H. pylori eradication was not confirmed by other authors (74,78,79). If considered the gut microbiome hypothesis, factors that can differentially modulate the gut microbial composition such as diet or the precise antibiotic combination used, might account for these discrepant results.
Associations between H. pylori infection and glucose homeostasis have been also explored (10,23,80,81). Evidence suggests that H. pylori may be involved in both diabetes onset and impaired glycemic control in diabetic patients, but contradictory trends or lack of association between H. pylori infection and type 2 Diabetes have been also reported (80,81). One of the mechanisms proposed for this relationship is the proinflammatory milieu induced by H. pylori infection that promotes gastric inflammation and cytokine secretion. In fact, H. pylori infection has been associated with increased levels of C-reactive protein (CRP) and interleukin 6 (IL6) (82,83).
The link between H. pylori and Type 2 Diabetes was first analyzed in 1989 by Simon et al., who found that the prevalence of H. pylori infection was higher in patients with Type 2 Diabetes compared to asymptomatic volunteers (84). Thenceforth, most of the studies have corroborated this finding reporting a higher prevalence of H. pylori in patients with Type 2 Diabetes (usually greater than 50%) in diabetic patients than in non-diabetic subjects (85)(86)(87)(88). Recently published meta-analyses this year aimed at elucidating the direction of the relationship between H. pylori and Type 2 Diabetes (81). Pooled H. pylori prevalence in diabetic patients was 54%, but there were high (even contradictory) regional variability as the highest prevalence of H. pylori infection in patients with T2DM was 66% in Africa and the lowest was 15% in USA (81). Moreover, it was found that patients with H. pylori infection had a higher risk of Type 2 Diabetes, but results again differed depending on geographical regions with a direct relationship in Europe, Asian and Africa, but a negative relationship in USA (80). In subgroup analysis, the relationship between H. pylori and the risk of diabetes was different according to age, level of glycated hemoglobin A (HbA1c), duration of diabetes and methods for H. pylori detection. This suggests that these factors could be an important source of heterogeneity in the studies included in the meta-analyses (80,81).
Furthermore, there are evidence suggesting that H. pylori infection is related to worse outcomes in diabetic patients. For instance, HbA1c levels were higher in diabetic patients with H. pylori infection in comparison with diabetic subjects negative for H. pylori (89). On the other hand, Yang et al. concluded that diabetic patients with H. pylori infection showed higher risk of cardiovascular diseases and more severe peripheral arterial stiffness than diabetic patients without H. pylori infection (90). Reciprocally, diabetic H. pylori positive patients showed worse symptomatology related to the infection (91,92). A higher degree of insulin resistance was also reported in H. pylori infected patients from different populations (8), an association that was confirmed in every studies included in a systematic review but one (88).
Therefore, although the link between H. pylori infection and Type 2 Diabetes has been controversial in some studies, the general trend suggests a higher susceptibility to H. pylori infection in diabetic patients. Likewise, H. pylori infection would promote the development of Type 2 Diabetes, as well as worse glycemic control and insulin responsiveness, likely due in part to the enhanced inflammation. In this line, gut microbiota could be a contributing factor as a modulator of the inflammatory response as well as of the secretion of incretins involved in glucose homeostasis as detailed in next section. In this regards, some clues have been given from intervention studies based on the use of a cocktail of antibiotics for H. pylori eradication and found that non-diabetic patients with H. pylori infection subjected to antibiotic treatment, experienced an improvement in insulin resistance index, HbA1c and insulin levels concomitant to a reduction of low-grade inflammation (10,87,93). Of note, efficiency on eradication rates of infection was lower in diabetic patients than in non-diabetic (8). Higher rates of antibiotic resistance in diabetic patients than in nondiabetic ones might account for these differences (94). Furthermore, the improvement in glucose homeostasis, i.e. HbA1c and glucose levels, in diabetic patients after successful eradication of H. pylori, were not statistically significant in comparison with non-diabetic controls or baseline values (95). Notably, gut microbiota composition varies in diabetic patients as compared to nondiabetic subjects (96)(97)(98).

GUT MICROBIOTA, ANTIBIOTIC THERAPY, AND METABOLIC DISEASES
The human gut is colonized by a myriad of microorganisms that encompass bacteria, fungi, archaea, protozoa and viruses. This complex ecological community comprises symbiotic, commensal and pathogenic microbes that physiologically interact with the host at different levels and functions. Locally, gut microbiota contributes to the maintenance of gut barrier function and integrity, fermentation of indigestible dietary substrates, vitamin synthesis and immune system regulation. However, the activity of these microorganisms goes beyond the gastrointestinal tract, affecting the function of distant organs including brain, liver, pancreas, muscle or adipose tissue (99).
The impact that gut microbiota has on host's metabolism is perhaps one of the most evident interrelations between the two systems. Besides harvesting energy from indigestible dietary fibers and modulating enterocyte function, gut bacterial community can also indirectly affect metabolic homeostasis by regulating immune system, local hormone secretion or bile acid synthesis. The efficiency of these actions depends on the specific and prevailing bacterial species that reside in the intestine (20,100). Several reports described differential bacterial profile in the gut from both obese and diabetic subjects (96)(97)(98). Although there are some controversial studies, in general terms, obesity-associated microbiota is depicted by an increase in Actinobacteria and Firmicutes and decline in Bacteroidetes phyla (97) as well as by diminished microbial richness and diversity (19,32,96,(101)(102)(103). Low microbial gene richness was associated with more adiposity, dyslipidemia and insulin resistance (101). In the same vein, diminished microbial gene richness has been related to lowgrade inflammation and marked dysmetabolism (102) and, even among individuals with severe obesity, those patients with lower microbial gene richness had worse metabolic conditions (104). Among microbial genes associated with obesity, 75% belongs to Actinobacteria and 25% to Firmicutes, whereas 42% of leanassociated microbial genes belong to Bacteroidetes (96,105,106). Gut microbiota composition from diabetic patients also differs from healthy subjects (107)(108)(109). Munukka et al., analyzed gut microbiota from women with and without metabolic disorders and concluded that the differences in some bacteria belonging to Eubacterium rectal-Clostridium coccoides group were more associated with obesity-related metabolic disorders than with obesity per se (110). Evidence also suggest that the amount of specific gut bacterial groups might also influence lipid metabolism.
To be more precise, Lachnospiraceae was related to low LDL-C levels and both Pasteurellaceae and Collinsella associated with low triglyceride levels. Tenericutes and Butyricimonas were also related to a favorable lipid profile with low triglyceride and high HDL-C levels (111). Therefore, gut microbiota has been related to the different hallmarks of the metabolic syndrome.
These data point out the gut microbiota as a crucial player in the regulation of host's energy homeostasis, though the mechanism by which the different bacterial species contribute to the regulation of metabolic functions are not fully understood. Current evidence point at low-grade inflammation and microbiota-derived metabolites as the possible links in the interplay of the gut microbiota and host's metabolism. These topics have been extensively reviewed elsewhere (19)(20)(21)112) and will only be briefly mentioned in this review.
As above mentioned, low-grade inflammation is a hallmark of obesity and its related metabolic complications. Inflammatory mediators interfere insulin signaling promoting insulin resistance. Immune cell infiltration in metabolic tissues impair its function and is enhanced in metabolic diseases. Concordantly, inflammation is related to diabetes, unhealthy lipid profile and favors the development of atherosclerotic lesions (113). However, the precise factors triggering the inflammatory response in metabolic diseases remains unclear. Within this context, it has been hypothesized that bacterial products, such as the Gram-negative bacteria cell wall component, lipopolysaccharide (LPS), can translocate from the gut to the circulation (114). LPS translocation would be enhanced by the consumption of fatty meals as these molecules are packed into chylomicrons in the enterocytes and delivered into the circulation (115). In addition, a "leaky" gut due to compromised gut barrier integrity would also favor LPS translocation (114). Circulating gutderived LPS, that activate the inflammatory response by binding to its receptor TLR-4, can reach metabolic tissues such as adipose tissue and induce inflammation, insulin resistance and impaired lipid accumulation (116). The inability of adipose tissue to store excess energy results in increased blood lipids and toxic lipid accumulation into non-fatty organs such as liver or muscle (117,118). In agreement with this hypothesis, circulating LPS levels are closely related to triglyceride levels and blood pressure, augment upon high-fat meal intake (115,116,119,120) and its increase is associated with inflammation and lower expression of lipogenic markers in the adipose tissue of obese patients, as recently confirmed by our group (116).
During the fermentation of indigestible dietary substrates, intestinal bacteria produces different by-products such as shortchain fatty acids (SCFAs). SCFAs, such as butyrate, acetate and propionate, take part in a variety of physiological functions. SCFAs have been shown to regulate adiposity, to improve insulin sensitivity, to exert anti-inflammatory action by modulating immune cells and to regulate incretin secretion. In addition, these metabolites also participate in the maintenance of the gut barrier integrity, decreasing the risk of bacterial product translocation from the gut to the circulation (19,21,(121)(122)(123). Butyrate represents a relevant energy source for colonocytes, contributes to epithelial cell health and consequently, to the integrity of intestinal epithelia. In addition, anti-inflammatory properties have been attributed to this SCFA (20,124). Propionate stimulates the release of the satietyinducing incretins GLP-1 and peptide YY, resulting in reduced food intake and concomitant weight, visceral adipose tissue and hepatic fat reduction, and preserved insulin sensitivity (125). Acetate is used by adipocytes, muscle or liver as energy substrate. It has been suggested that elevated levels of acetate and propionate associates with satiety, weight loss, and decreased inflammatory markers and blood lipids (126). However, some controversial results indicate that, when highly produced by gut microbiota, acetate could activates several pathways stimulating insulin secretion, hyperphagia and obesity in animal models (127). Then, the harmful or beneficial effects of SCFAs might be dependent on their levels as well as on the interaction with other factors (20). Notwithstanding this, human studies showed that obese and diabetic patients had diminished SCFA-producing bacteria and SCFAs (128). Therefore, SCFAs can mediate the multiple effect that gut microbiota exert on host´s metabolism.
Modulation of bile acid dynamics by gut microbiota has been also related to metabolic regulation (21). Bile acids participates in fat emulsification and absorption of lipids and liposoluble vitamins in the intestine. Primary bile acids are produced by hepatocytes, released in the proximal intestine, reabsorbed in distal ileum and recycled by the liver. Besides facilitating fat digestion and absorption, bile acids exerts metabolic functions by means of their receptors farnesoid X receptor (FXR) and Takeda G-protein receptor-5 (TGR5). FXR and TGR5 signaling promotes hepatic glycogen synthesis and insulin sensitivity, insulin secretion, energy expenditure in liver brown adipose tissue and muscle, thermogenesis and satiety (76). Gut microbiota metabolizes and deconjugates primary bile acids to transform them into secondary bile acids (129). Therefore, gut microbiota play a relevant role in determining the composition of bile acid pool which is relevant to the final biological actions of these molecules (130). In fact, it has been suggested that secondary bile acids can differentially interact with FXR and TGR5 than primary bile acid which would result in distinct metabolic actions (112,131).
Obviously, bacterium by-products generated by the gut microbiota is highly determined by the diet. Bacteria are highly specific on the kind of nutrient they metabolize. Then, the effect of a certain type of diet on adiposity and metabolism activation not only is determined by calorie amount or energy harvesting by gut bacteria, but it is also defined by the specific kind of nutrient that enhances the growth of specific bacterial species damaging to coexisting microorganism by bacterial competition (99). For instance, some gut bacterial groups metabolized choline and L-Carnitine from dietary source (e.g. red meat or eggs) into trimethylamine (TMA). TMA is absorbed and oxidized by hepatic flavin monooxygenase 3 (FMO3) to produce trimethylamine N-oxide (TMAO) (132). Elevated levels of this metabolite have been described in diabetic (132) and obese (133) individuals and correlates with cardiovascular disease (134).
The production of these metabolites, the integrity of gut mucosa and consequently the actions that gut bacteria would exert as a whole, would depend on the abundance and proportion of each bacterial taxa inhabiting our intestine. Determinants of gut microbiota composition include dietary intake or antibiotic treatment, among other exogenous and endogenous factors (20,30,135,136). Disruption of the gut bacterial equilibrium results in the deregulation of host metabolic functions and triggers the onset of obesity and obesity-related metabolic diseases (100,112).
In this line, several studies aimed at assessing whether the modulation of the gut microbiota by antibiotic administration was related to obesity and other metabolic variables (31-34, 137, 138). Associations of diverse antibiotic treatments in early life with increased risk of childhood obesity has been reported (33,34). A recent study focused on analyzing the gut microbiota profiles associated with the increased risk of childhood obesity due to early antibiotic exposure. This study showed that boys were at higher risk of increased abdominal adiposity than girls exposed to several courses of antibiotics and that changes in specific bacterial groups were related to both repeated antibiotic exposure and childhood adiposity (31). By contrast, Ajslev et al., described that, when mothers were overweight, early antibiotic exposure were related to decreased risk of developing childhood obesity. Unfortunately, gut microbiota was not analyzed in this study so it cannot be determined whether specific modifications in gut bacterial groups could account for the differential risk of developing obesity depending on maternal weight (137).
In the adult population, the association between antibiotic treatment, gut microbiota and insulin sensitivity is still controversial (32,138). Vrieze et al., found that 1-week vancomycin treatment increased and decreased primary and secondary bile acids, respectively, and decreased insulin sensitivity evaluated 2-3 days after treatment cessation. However, amoxicillin administration had no effects on the study variables (32). Despite significant changes in gut microbiota diversity and composition as well as in SCFA and bile acids levels, Reijnders et al., did not find significant effects after 1-week vancomycin or amoxicillin exposure on different markers of insulin sensitivity, metabolic, hormonal or inflammatory parameters evaluated 8weeks after treatment (138).
These findings point out the fact that factors such as age, gender, previous obesity degree, the type of antibiotic or posttherapy evaluation time could influence the effects on host's metabolism. Therefore, further studies are required to elucidate how specific gut microbiota modulation exerted by each kind of antibiotic therapy affects metabolic homeostasis according to host's characteristics. In view of the multiple therapeutic options for H. pylori eradication, this perspective become particularly relevant in the management of H. pylori infection. In addition, it is of interest to elucidate whether gut microbiota modulation by antibiotic eradication therapy improves or worsens the previous impaired metabolic status of H. pylori patients.

GUT MICROBIOTA IN H. PYLORI INFECTION
H. pylori colonizes gastric mucosa leading to modification in the gastric microenvironment and disturbing gastric microbiota composition (139). These changes has serious local effects on stomach, but it also affects the function of other regions of the gastrointestinal tract which likely results in altered absorption of nutrients and drugs as well as on the production of incretins involved in metabolic homeostasis. Furthermore, few years ago, it was also proposed that H. pylori infection might also trigger large intestinal microbiota changes (140). Although close attention has been paid to the effect of H. pylori eradication treatment on gut microbiota, very recent human studies have also confirmed that H. pylori infection per se is linked to gut dysbiosis including alterations in bacterial diversity and abundance (9)(10)(11)(12)(13)(14)(15)(16)(17)(18). A summary of bacterial shifts reported in individuals infected by H. pylori is displayed in Table 1.
Within this context, our group and others have reported that human H. pylori infection is related to changes in bacterial diversity, but findings still controversial (10)(11)(12)(13)(14)16). Some studies observed less diversity in H. pylori positive than in H. pylori negative subjects (10,11). In view of the fact that H. pylori infection is usually related to unhealthy profile, these results are congruent with the general idea that high gut microbial diversity is an indicator of a healthy gut microbiome (101). In fact, low bacterial richness has been related to insulin resistance, dyslipidemia and higher overall adiposity (32). Nevertheless, other authors intriguingly found that several alpha diversity estimators exhibited higher scores, which indicate higher level of gut microbiota complexity in H. pylori cases (12)(13)(14)16). Different study population, calculation of different diversity indexes or techniques for quantifying gut microbiota might underlie this opposite findings. In addition, other factors such diet or disease severity could be also influencing gut diversity. Therefore, controlled trials with homogenous designs are required to elucidate the actual effect of H. pylori infection and diversity of gut microbiota.
Regarding the specific groups of gut bacteria, some human studies have shown that the relative abundances of dominant phyla Bacteroidetes, Firmicutes and Proteobacteria significantly differ in the gut of H. pylori positive individuals compared to negative controls, although trends towards the increase, decrease or no differences have been reported (10,(14)(15)(16). Similarly, some studies found that H. pylori infection decreases the abundance of Actinobacteria (10, 11), but He et al., reported an increase in this Phylum (16). Similar to bacterial diversity, variability in study population including age or geographical region and different exclusion and inclusion criteria might underlie the heterogeneity in major phyla proportion in H. pylori infected patients. In fact, it has been reported that gut microbiota can differ between Japanese and other populations (142). Differences in these phyla have been also related to obesity and Type 2 Diabetes (19). Microbial pattern with low Bacteroidetes and high Actinobacteria and Firmicutes proportions characterized the gut microbiota from obese patients (19,96,103). Therefore, further studies on H. pylori positive patients with more homogenous population might be helpful to establish consistent trends among predominant phyla in relation to metabolic diseases.
Despite not having found consistent results in the main phyla related to H. pylori infection, in general terms and taking into account the diversity of bacterial functions included in each phylum, bacterial changes at the level of the different taxonomic categories belonging to each phylum may have both positive and negative effects on the host. In this line, Frost et al., found that H. pylori infected individuals displayed elevated levels of the facultative pathogen Haemophilus but decreased levels of Pseudoflavonifractor and Parasutterella (13). Pseudoflavonifractor encompasses butyrate-producing bacteria and Parasutterella are succinate-producing bacteria (143). Thus, the decrease of Pseudoflavonifractor might partly explain the negative impact of H. pylori infection on metabolic profile as the result of reduced SCFA production. Notwithstanding this, the genus Parasutterella is increased in Crohn's disease (144), so authors hypothesized that H. pylori might be exerting a protective role against pathogens related to other gastrointestinal diseases (13). In fact, succinate has been described as a virulence factor that might exacerbate enteric infections (145). In addition, this bacterium together with other taxa increased by H. pylori (e.g. Alistipes), have been proposed as potential predictor biomarkers of obesity-related metabolic abnormalities (146). In the same vein, succinate-producing bacteria and elevated levels of succinate have also been associated with Type 2 Diabetes and obesity (147).
By using shotgun metagenomic sequencing, Wang et al., found variation in other bacterial species according to the presence or absence of H. pylori infection (141). They specifically found increased proportions of Prevotella copri, a proinflammatory bacterium (148) as well as Klebsiella pneumoniae and Enterobacter cloacae, two infectious bacteria (149). These results suggest that H. pylori infection may promote growth of harmful bacteria in the gut. Conversely, Sutterella wadsworthensis, B. vulgatus, and E. coli amounts were lower in the H. pylori-positive compared to noninfected individuals (141). Members of the genus Sutterella are highly prevalent commensals with the ability to adhere to intestinal epithelial cells indicating a possible immunomodulatory role (150). In addition, the pathobiont Bacteroides vulgatus has been implicated in the etiology of both Crohn's disease and ulcerative colitis, but little is known about how its functional activity might drive the host inflammatory response (151).
A study carried out in a Japanese population specifically focused on analyzing how H. pylori modulates the proportion of each Lactobacillus species in the gut and found higher relative abundance of Lactobacillus in H. pylori infected patients than in non-infected individuals (18). Lactobacilli endows gastrointestinal tract with "colonization resistance" serving as a defense mechanism which protects the host from potentially pathogenic microbials. Probiotic effects have widely been attributed to different Lactobacillus species such as L. casei or L. rhamnosus that prevented antibiotic-related diarrhea (152), which may be due to the role of these bacteria in gut barrier preservation (153). Furthermore, certain Lactobacillus species have been also related to positive metabolic outcomes. While Lactobacilli relative abundance in the gut microbiota in obesity and after weight loss is controversial, it has been described that specific Lactobacillus strains (e.g. L. rhamnosus, L. gasseri, L. plantarum or L. paracasei) relates with a reduction of obesity-associated metabolic disorders and even an improvement on insulin sensitivity. However, other species such as L. reuteri has positively been associated with adiposity, which suggests Lactobacillus strain-dependent physiological effects on metabolism and weight regulation (19,(154)(155)(156). Iino et al., found that H. pylori positive patients displayed reduced amounts of L. acidophilus and increased proportion of L. salivarius in comparison with noninfected subjects. Authors suggested that alterations in Lactobacillus proportions could be related to the suppression of gastric acid secretion by H. pylori infection, but putative physiological effects that these changes can have on host's metabolism and health have not yet been addressed (18).
In the same way that gastric colonization by H. pylori can affect the composition of gut microbiota, some gut bacteria might be also influencing the bacterial colonization of other gastrointestinal regions, including H. pylori in the stomach. In this regard, Chen et al., found that, despite not having noticing significant differences in the abundance of several putative beneficial taxa including Bifidobacterium, Lactobacillus, Clostridium butyricum, Faecalibacterium prausnitzii and Akkermansia muciniphila, the phylum Nitrospirae exclusively appeared in H. pylori-negative subjects (17). Similarly, in the study by Wu et al., this phylum presented low values in patients with duodenal ulcer and H. pylori infection (11). Nitrospirae are the most abundant and diverse nitriteoxidizing bacteria which convert nitrite to nitrate (157). Of note, nitrite was shown to have bactericidal effects against H. pylori (158).
Differences between the gut microbiota from asymptomatic H. pylori infected patients vs. non-infected subjects have been also reported (12,15,16). In fact, H. pylori infection alters the gut microbiota in this asymptomatic patients by increasing Proteobacteria, Clostridium, Firmicutes and Prevotella in a paediatric population (15) and members belonging to Succinivibrio, Coriobacteriaceae, Enterococcaceae, and Rikenellaceae in adults (12) compared to non- infected subjects. The study by He et al., identified their study participants as asymptomatic although they also stated that subjects were diagnosed with superficial gastritis (16). To the best of our knowledge, it has not yet been evaluated whether the gut microbiota from asymptomatic H. pylori-infected patients also differs from that from symptomatic H. pylori-infected patients. This information could be of great interest to shed light on the possible role of the gut microbiota composition as player and/or biomarker of the severity of H. pylori infection which ranges from asymptomatic to serious clinical manifestation such as gastric cancers.
In this line of thought, Gao et al. investigated the association between the gut microbiota and H. pylori-related gastric lesions. To this end, they compared the intestinal microbiota of subjects without gastric injury, with gastritis or with intestinal metaplasia, and concluded that indeed the alterations of the fecal microbiota, particularly the phyla Bacteroidetes, Firmicutes and Proteobacteria, are likely involved in the process of progression of the gastric disease associated with H. pylori (14).
In addition, Frost et al., found that there were bacterial significant variations even within the H. pylori positive patient group that was dependent on the individual H. pylori antigen load. To be more precise, a negative association was reported between antigen load and Bacteroides, Fusicatenibacter, Barnesiella, and Alistipes, taxa exhibiting putative healthy characteristics at first sight. Fusicatenibacter and Alistipes are butyrate and lactate producing bacteria that could exert anti-inflammatory and other beneficial actions (159,160). However, opposite results regarding different species belonging to Alistipes taxa and its relationship with dysfunction of several organs such as liver or cardiovascular system has been described (161). In fact, He et al., who also found that H. pylori infected patients had elevated levels of Alistipes compared to non-infected patients, referred to an animal study which found that this taxa was positively correlated with weight, fat mass, LPS and inflammatory cytokines (16,162). On the other hand, Barnesiella is associated with more efficient eradication of antibiotic-resistant bacteria (163). By contrast, authors stated that non-significant results were retrieved from analyses taking into account H. pylori serology level (13). In the same vein, Iino et al., found that the relative abundance of Lactobacillus was significantly higher in H. pylori positive patients with severe atrophic gastritis as compared with infected patients with mild atrophy gastritis or without gastritis (18) supporting the hypothesis that gut microbiota could also differ depending on symptoms severity. Of note, no differences according to the degree of atrophic gastritis was found in the specific Lactobacillus species studied (L. acidophilus and L. salivarius) (18) which remarks the necessity to find the specific bacterial groups that can be related to infectious severity.

MODULATION OF THE GUT MICROBIOTA BY ANTIBIOTIC-BASED H. PYLORI ERADICATION THERAPIES
According to the Kyoto consensus report, H. pylori gastritis has to be regarded as an infectious disease and the resulting recommendation is the treatment of all H. pylori infected patients regardless the clinical manifestations (164). Antibiotic therapies are the first-line treatment for H. pylori eradication, but the increasing antibiotic resistance rates make necessary the adaptation of the treatment regimen. In general terms, the standard triple therapy based on the combination of one proton pump inhibitor (PPI), clarithromycin and amoxicillin or metronidazole is recommended when H. pylori clarithromycin resistance is low (<15%). By contrast, quadruple therapy based on bismuth administration together two antibiotics and one PPI is the first-line treatment for H. pylori eradication in regions with high antibiotic resistance rates (27,28). However, recent reports highlight how the increasing resistance to antibiotics compromises the efficacy of these recommended therapies and salvage regimens have to be used (28,29). This lead to high variability on the type and amount of specific antibiotics that can have different impact on the patient's health.
Antibiotic H. pylori therapies are not exempt from side effects. The extensively reported gastrointestinal adverse events associated to antibiotic administration such as nausea, vomiting or diarrhea among others, are linked to quantitative and qualitative changes in the gut microbiota (165). Antibiotic-induced changes in the gut microbiota has become a matter of concern in the treatment of H. pylori as the essential role of intestinal-residing bacteria in maintaining human health beyond gut function is increasingly accepted (27).
In general terms, antibiotic administration has been related to a decrease in bacterial diversity (30,138). Regarding the various H. pylori eradication therapies it has been reported a decrease in bacterial diversity after triple therapy consisting of PPI, amoxicillin and clarithromycin for 7 days (39,40,43), 10 days (10) or 14 days (11,36). Nevertheless, most of these studies found that bacterial diversity was restored in the short and the long-term after treatment cessation (11,36,40,43). However, other studies did not reported an improvement of bacterial diversity after cessation of triple therapy (10,39).
Quadruple therapy with bismuth during 10 or 14 days also led to decreased bacterial diversity but not all the studies found a total recovery of bacterial diversity evaluated from 6 weeks to 1 year posttreatment (16,36,38,44). By contrast, concomitant therapy of three antibiotics (amoxicillin, clarithromycin, metronidazole) without bismuth for 14 days decreased bacterial diversity, but it was fully (35) or partly restored (36) at 2 months or 1 year posttreatment, respectively.
Liou et al., compared the effect of triple, quadruple and concomitant therapies on bacterial diversity and found that the two latter treatments were related to lower alpha diversity in comparison with triple therapy at 1 year post-eradication treatment (36).
Notably, few studies have contrasted eradication therapyinduced bacterial diversity changes with bacterial diversity in the gut of non-infected patients (controls) which would drop hint at the similarity of bacterial diversity scores before and after  eradication treatment to those seen in healthy conditions. However, opposite evidence has been found with no differences in the gut bacterial diversity at 6 and 26 weeks post-treatment with quadruple therapy compared to controls (16) or lower bacterial diversity evaluated at 2 months after triple therapy than in both controls and H. pylori infected patients before undergoing eradication intervention (10). Pooled together, these findings suggest that the effect on gut bacterial diversity and its length are dependent on the eradication therapy used. A recent meta-analysis stated that alpha diversity was reduced immediately within the first week after eradication, but no consistent conclusions were drawn from studies evaluating bacterial diversity at longer evaluation times (168). It is worth of mention that this meta-analysis pooled different kinds of eradication therapies to analyze the effect on bacterial diversity, which could limiting the conclusions by adding variability to the results. This meta-analysis conclude that further studies will be required to gain more evidence before raising firm conclusions.
Most of the studies tested whether antibiotic-induced changes in the abundance of bacterial groups were restored upon treatment cessation. As summarized in Table 2, gut microbiota composition is restored in most cases at 2 months post-treatment. Nevertheless, it has been also reported that the imbalance of some bacterial groups remains in the short-and the long-term after treatment cessation (10,16,24,40,42). Jakobsson et al., documented changes in the gut microbiota that persisted for up to 4 years after H. pylori eradication, but formal statistical analysis was not done owing to small sample size (40). In the same way, other studies also reported persistent changes in gut microbiota i.e. a decrease in Proteobacteria, Bacterioidetes, Actinobacteria and an increase in Firmicutes at 6 (16, 42) and 12 moths post-triple therapy (42), as well as an enrichment in Proteobacteria, Actinobacteria, Verrucomicrobia compared to baseline values at 18 months after treatment (42). Elevated relative Bacteroidetes : Firmicutes ratio was observed at 2 and 3 months of treatment cessation (10,24). At phylum level, the meta-analysis made by Ye et al. showed a reduction in Actinobacteria and Bacteroidetes during the follow-up. In addition, Firmicutes was exclusively found to be augmented in the long-term after H. pylori eradication while Poteobacteria increased in short-term and returned to normality in the long-term (168). Lasting alterations in the proportions of these phyla may result in an altered production of bacterial metabolites which would affect hostbacterial crosstalk. For instance, acetate and propionate are mainly produced by Bacteroidetes members, while Firmicutes members typically produce butyrate (171). Notably, reduced Bacteroidetes: Firmicutes ratio and increased abundance of Proteobacteria, have been associated with obesity and the metabolic syndrome (19,172).
Persistent antibiotic-induced shifts on bacterial groups can be associated with therapy adverse effects. In agreement with this, the relative abundance of the phylum Proteobacteria and some of its members including Aggregatibacter and Sutterella were higher at 2 weeks after quadruple therapy in patients who suffered from eradication intervention side effects compared with patients that did not reported adverse symptoms (44). Proteobacteria phylum includes many pathogens and it has been proposed that may be partly responsible for the development of adverse effects during eradication therapy (44).
Interestingly, some studies assessed the effect of probiotic supplementation administered together the antibiotic combination which attenuated the antibiotic-induced imbalance on gut microbiota composition (11,41). Furthermore, probiotic supplementation have been associated with improved gastrointestinal symptoms (17,39) and increased Bacteroidetes: Firmicutes ratio (17). All together, these effect might help to build up a beneficial gut microbiota profile after eradication therapy (38). This suggest that probiotic administration could attenuate antibiotic-induced gut dysbiosis, but to the best of our knowledge, it remains unexplored the consequences that this can have on host's metabolic health within the context of H. pylori infection.
The discrepant observations between studies could be due to different eradication regimens, drug doses, treatment duration and/or sample size. Furthermore, other factors such as dietary habits, resistance to antibiotics and differences in the rate of absorption of antibiotics can affect the influence of eradication therapy on the intestinal microbiota and the time required to restore original bacterial composition. On the other hand, it is remarkable that most of the studies did not include controls without H. pylori infection, and considered that the gut microbiota was restored when no significant differences were observed regarding baseline. In this sense, it should be taken into account that pre-treatment bacterial composition is already influenced by H. pylori infection. This fact makes difficult to determine whether bacterial communities post-therapy mirror a healthy gut microbiota. All in all, further studies analyzing the effect of each eradication therapy and including non-infected controls, are necessary to clarify the effects that the various therapies for H. pylori eradication have on gut microbiota in the long-term and to elucidate the factors responsible for the variability in antibiotic response.

RELATIONSHIP BETWEEN METABOLIC TRAITS AND GUT MICROBIOTA MODIFICATIONS INDUCED BY H. PYLORI INFECTION AND THERAPY
As detailed above, evidence has shown that H. pylori infection as well as its eradication treatment lead to gut dysbiosis. However, despite the fact that gut microbiota is closely linked to metabolic health, the role of gut microbiota in the relationship between H. pylori infection and metabolic dysregulation has been scarcely studied. Several studies from our group, which analyzed gut microbiota from H. pylori infected patients before and after eradication treatment (antibiotic triple therapy based on omeprazole, clarithromycin, amoxicillin) as well as noninfected control patients, addressed this issue (Table 3) (10,22,23). It was found that H. pylori eradication treatment produces specific bacterial shifts associated with changes in glucose homeostasis-related parameters [HbA1c, glucose area under the curve (AUC) calculated upon an oral glucose tolerance test (OGTT) (10)], GLP-1 (23), ghrelin (24) and HDL-C levels (22).
To be more specific, changes in the amount of Rikenellaceae, Butyricimonas, E. biforme, B. fragilis, and Megamonas were inversely associated with changes in glucose levels or glucoserelated parameters, i.e. HbA1c, in H. pylori subjects after eradication treatment (10). Several studies have shown that these bacteria are involved in the fermentation of nondigestible carbohydrates and the generation of SCFAs such as acetate, propionate and butyrate (173)(174)(175). In addition, Rikenellaceae and Butyricimonas members are also able to use the environmental glucose to produce SCFAs, which could contribute to glucose level regulation (175,176).
It was not only analyzed glucose metabolism, but also the dynamics of GLP-1 in the presence of H. pylori and upon antibiotic therapy (23). GLP-1, mainly produced in the ileum and colon by enteroendocrine L cells, regulates glucose homeostasis by systemic effects on pancreatic b cells, reduction of gastric acid secretion, delaying gastric emptying, regulating appetite as well as adipose tissue physiology (177,178). Taking into account that some gut bacteria modulate intestinal enteroendocrine L cell secretion of GLP-1 (122,179), it is of interest to explore whether both H. pylori-induced or antibiotic treatment-induced modifications in gut microbiota could also influence GLP-1 levels. In this way, GLP-1 levels at different time point after an OGTT before and after eradication treatment in H. pylori positive patients as well as in H. pylori negative controls were evaluated. We found that changes in GLP-1 AUC after eradication therapy positively and negatively correlated with the changes in the genus Lachnobacterium and Bifidobacterium adolescentis, respectively. Variation in GLP-1 at 60 min after OGTT and the changes in the family Coriobacteriaceae also positively correlated two months after the treatment (23). Species from the genus Lachnobacterium and Bifidobacterium are able to produce SCFAs (180). Among the beneficial effects attributed to SCFAs, these metabolites have also been proposed to favor GLP-1 L cell secretion and to exert anti-inflammatory action which may have beneficial effects on glucose homeostasis and insulin sensitivity (181,182). Studies analyzing the use of Bifidobacterium as probiotic reported an increase in GLP-1 production and beneficial effects on carbohydrate metabolism (183). Concordantly, type 2 diabetic patients harbored smaller amounts of B. adolescentis (184). The cause and consequences of these opposite correlations between these two groups of gut bacteria and GLP-1 secretion in H. pylori positive patients after eradication treatment should be further analyzed. Bacteria belonging to the family Coriobacteriaceae, which has been shown to be reduced in Type 2 diabetic women (185), might also indirectly favors GLP-1 secretion by generating bile acids that stimulates GLP-1 secretion in L cells (19,186). Relationships between microbial groups and GLP-1 levels differed between H. pylori infected patients and non-infected controls. Positive correlations between the genus Megamonas [previously related to carbohydrate metabolism (187)] and GLP-1 levels were exclusively found in the control group. By contrast, bacteria belonging to the phylum Proteobacteria correlated positively with GLP-1 levels only in H. pylori-positive patients. Notably, these correlations disappeared after the eradication treatment and shifted to Bifidobacterium longun and the genus Prevotella which correlated positively and negatively with GLP-1 levels, respectively (23).
Other authors also explored the influence of H. pylori eradication therapy-related gut microbial changes on ghrelin levels (24). Ghrelin is a multifunctional hormone mainly secreted by gastric mucosa that regulates body weight by stimulating appetite, growth hormone secretion, fat storage among other relevant systemic functions in energy metabolism (188). Yanagi et al., reported that changes in the Bacteroidetes:Firmicutes ratio were inversely related to changes in plasma ghrelin levels after the administration of triple therapy for H. pylori eradication (24). Paradoxically, previous studies showed that modulation of the gut microbiota by using prebiotic supplementation reduced ghrelin secretion (123).
The relationship between specific profiles of gut bacteria in H. pylori infected patients before and after eradication treatment and metabolism are not restricted to carbohydrate metabolism but it seems to extend to lipid metabolism. Major phyla Bacteroidetes (increased in H. pylori positive patients) and Firmicutes (decreased in H. pylori positive patients) were negatively and positively correlated with HDL/LDL ratio, respectively. When analyzed in more detail, several bacterium taxa (Eubacterium, Bacteroides coprophilus, E. biforme), that were increased in H. pylori positive patients, were related to HDL/LDL ratio (22). Specific associations with lipid profile related to eradication treatment-induced gut microbial modifications were also found. Positive and negative changes in Delsufovibrio and Rikenellaceae, respectively, predicted changes in HDL-C levels at month 2 after completing antibiotic treatment (22). Delsufovibrio is a producer of LPS (189), but its products can also up-regulate the expression of the critical regulator of lipid absorption, CD36 (190), that has been positively associated with HDL-C levels (191). On the other hand, Rikenellaceae produces acetate that, besides the beneficial SCFA effects, has been described to promote hepatic De novo lipogenesis and cholesterol synthesis (192).
All in all, this emerging evidence suggest that gut microbiota shifts induced by H. pylori and upon antibiotic treatment for its eradication might, at least in part, underlie modulation of metabolic variables. However, further studies should be performed to confirm this hypothesis and to fill the many gaps of this intricate cross-regulation. Other H. pylori therapies than triple therapy should be assessed, as well as the concomitant effect of pre-or probiotic supplementation during eradication treatment. In addition, the precise mechanisms and bacterial activities involved in the H. pylori-gut microbiota-metabolism crosstalk remains to be elucidated.

DISCUSSION
While the number of studies analyzing the effect of H. pylori eradication therapies on the gut microbiota are increasing in recent years, current evidence is not enough to draw clear conclusions on the most relevant and common bacterial shifts. Future research needs to face several challenges to make clear assumptions on this relationship. Different population characteristics can lead to divergent results. Gut microbiota composition varies among ethnic groups, likely due to the different dietary, hygienic and genetic factors as well as regional antibiotic resistance rates (142). Thus, the modifications on specific bacterial taxa and the underlying related mechanisms involved in metabolic regulation could largely differs depending on the region where the study is performed. In this sense, the inclusion of noninfected healthy individuals as controls in intervention studies would give valuable information to determine the actual reversion degree to "healthy" baseline gut microbiome. Moreover, it is likely that the severity of the H. pylori eradication therapy impact on metabolism and adiposity in the long-term also varies depending on the age group. In that way, other antibiotic treatments than those used for H. pylori eradication have been shown to influence the onset of obesity when applied to early-life infants, while evidence on antibiotic administration effect on adult metabolism is not so clear (31-34, 137, 138). The fact that most of the studies analyzing the gut microbiota in H. pylori infected patients excluded subjects with diabetes, obesity and cardiovascular diseases (10,16,17) also limits the knowledge on the concomitant management of the dysbiosis induced by H. pylori infection and treatment in patients with these metabolic diseases.
Other unaddressed issue is how the different intestinal bacterial patterns influenced by infection severity relates to metabolic homeostasis. Furthermore, the relationship between H. pylori-induced changes in gut microbiome and metabolic disorders has been scarcely explored as yet, and the effect of each therapeutic options used to treat H. pylori infection as well as the concomitant use of probiotics on specific metagenomic and metabolic modifications should be also further addressed.
Finally, to the best of our knowledge, direct evidence on the bacterial functions and actions (by-product production, inflammation and translocation of bacterial products) resulting from the modulation of the diverse bacterial groups as a whole during H. pylori infection and treatment remains unexplored. This might be helpful to improving the understanding of the metabolic regulation by gut microbiota within the context of H. pylori-related disease.
In conclusion, although current data points at an essential role of gut microbiota as mediator of the crosstalk between H. pylori and host's metabolic health, the numerous remaining unanswered questions warrant future in-depth research in this field.

AUTHOR CONTRIBUTIONS
MC-P and FJT contributed to conception and design of the manuscript. GMM-N, IC-P and MC-P researched the literature. GMM-N, IC-P, MC-P, and FJT drafted and wrote different sections of the manuscript and contributed to the discussion and interpretation of data. All authors contributed to the article and approved the submitted version. FUNDING GMM-N was supported by a Juan de la Cierva, Formació n contract (FJCI-2017-34349; Ministerio de Ciencia, Innovació n y Universidades; Spain). MC-P was recipient of a postdoctoral grant Juan de la Cierva Formació n (FJCI-2017-32194) from the Ministerio de Ciencia, Innovació n y Universidades (Spain) and a postdoctoral research grant (DOC_00448) from the Consejeria de Economí a, Industria, Conocimiento y Universidades (PAIDI, Junta de Andalucí a, Spain), co-funded by the Fondo Europeo de Desarrollo Regional (FEDER). IC-P has been recipient of postdoctoral grants Rí o Hortega (CM 17/00169) and Juan Rodé s (JR19/00054) from the Instituto de Salud Carlos III (ISCIII) and co-funded by Fondo Europeo de Desarrollo Regional (FEDER). This research was funded by Centros de Investigació n Biomé dica en Red (CIBER, CB06/03/0018) from the ISCIII, (Madrid, Spain); RIC-0539-2018 and PI-0092-2017 from Consejerí a de Salud (Junta de Andalucí a, Spain); PI18/ 01160 from the ISCIII (Madrid, Spain), and co-funded by the Fondo Europeo de Desarrollo Regional (FEDER).