Recent studies have suggested that gut-brain axis may be one of the mechanisms of major depression disorder (MDD). In animal studies, alteration of gut microbiota can affect animal's depression or anxiety-like behavior, brain neurochemistry and inflammation. In human studies, the composition of gut microbiota is different between patients with MDD and healthy controls. In addition, supplementation of probiotics can improve mood status in community and clinical participants. Inflammation is one of possible pathway to connect gut and brain. Gut permeability and inflammation level are higher in patients with MDD. Lactobacillus plantarum PS128 in one of bacteria extracted from traditional fermented food, Fu-Tsai. It can alleviate depressive-like behavior reduce inflammation level in maternal separation mice. This study is an 8-week open trial to investigate the effects of Lactobacillus plantarum PS128 on psychophysiology in patients with MDD and higher level of inflammation.

This is a two-phase study. In the first phase, we will recruited patients fulfilling the following inclusion criteria: Age 20-65; fulfill Diagnostic and Statistical Manual of Mental Disorders fifth version (DSM-V) criteria of major depressive episode in recent 2 years; Psychotropics including antidepressants, antipsychotics and hypnotics have been kept unchanged for at least 3 months. The exclusion criteria are: comorbid with schizophrenia, bipolar disorder, or other substance use (except tobacco) disorder; having active suicidal or homicidal ideation; known allergy to probiotics; comorbid with hypertension, diabetes mellitus, irritable bowel syndrome, inflammatory bowl disease, liver cirrhosis, or autoimmune diseases; known active bacterial, fungal, or viral infections in one month; use of antibiotics, steroid, immunosuppressants, probiotics, or synbiotics in the month before collecting blood and fecal samples; pregnant or lactating women; who state to have dietary pattern changed or in diet within previous two months. Those hs-CRP > 3 mg/L in the first screen will be invited into the second phase intervention. In the second phase intervention, we will give eligible patients Lactobacillus plantarum PS128 for 8 weeks, and compare depression symptoms, gut microbiota, gut inflammation and permeability, and serum inflammation level before and after intervention.

1. Background Major depressive disorder (MDD) is a complex and long-term illness that involves marked disabilities in affected patients. It has been one of the leading causes of global burden of disease, especially in the middle-age groups (GBD Lancet 2016; 388:1603-58). It is also a common comorbidity in patients with irritable bowel syndrome, cardiovascular disease and cancer (Musselman et al. 1998; Spiegel et al. 2003; Whitehead et al. 2002). The mechanisms responsible for depression and its comorbidity are not clear, but inflammatory processes may be implicated. Meta-analysis has shown that patients with depression have higher levels of inflammatory markers, such as C-Reactive Protein (CRP), interleukin-1 (IL-1), interleukin-6 (IL-6), and tumor necrosis factor (TNF)-alpha (Dowlati et al. 2010; Howren et al. 2009). Microbiota-gut-brain axis may explain higher inflammatory cytokines in patients with MDD. Accumulating data now indicate that the gut microbiota can communicate with the CNS--possibly through immune, neural, and endocrine pathways--and thereby influences brain function and behavior (Cryan et al. 2012; Forsythe et al. 2016; Rogers et al. 2016). Probiotics, which are defined as a live organism that when ingested in adequate amount exerts a health benefit, can affect gut microbiota, brain neurochemistry, HPA axis, and inflammation, and may be a novel class of psychotropics (Dinan et al. 2013).

1.1 Microbiota-gut-brain axis The lower gastrointestinal tract contains almost 100 trillion microorganisms, most of which are bacteria. The microbiota collectively encodes more than 3.3 million non-redundant genes, exceeding the number encoded by the human host genome by 150-fold (Qin et al. 2010). Many microbial gene products have important effects on metabolism and the health of the host.

There have been numerous studies supporting the communications between gut microbiota and brain. Study using germ free (GF) animal is a good model to show the relationship between microbiota and brain. Compared to specific pathogen free (SPF) mice with a normal gut microbiota, GF mice display increased motor activity and reduced anxiety behavior. An increased turnover of dopaminergic and serotonergic neurotransmitters in striatum was observed in GF mice (Diaz Heijtz et al. 2011). In addition, GF mice exhibited higher adrenocorticotropic hormone (ACTH) and corticosterone elevation in response to restraint stimulus than SPF mice did (Sudo et al. 2004). It suggests that gut microbiota can affect host's response to stress.

Several studies have shown that manipulation of gut microbiota can affect the presentation of host's psychophysiology. For example, consumption of high fat diet (HFD) is associated with altered microbial diversity (Daniel et al. 2014) and reduced synaptic plasticity (Liu et al. 2015). HFD also increased anxiety-like behavior (Sharma et al. 2013). Adult male C57BL/6 mice given HFD microbiota had significant and selective disruptions in exploratory, cognitive, and stereotypical behavior compared with mice with control diet microbiota. No significant differences in body weight were noted. The difference in composition of gut microbiota between HFD and normal-diet groups was noted (Bruce-Keller et al. 2015).

Administration of antibiotics is another way to affect gut microbiota composition and change host's psychophysiology. Oral administration of non-absorbable antimicrobials transiently altered gut microbiota composition and increased exploratory behavior and hippocampal expression of BDNF mRNA in mice, but intraperitoneal administration had no effect on behavior (Bercik et al. 2011). Increased antibiotic exposures were associated with the risk of depression and anxiety in a population study (Lurie et al. 2015). Therefore, administration of probiotics may be a direct way to change the environment and composition of gut microbiota, and observe its physiological effects on host.

1.2 Administration of Probiotics in animal studies Animal studies have shown that probiotics affect psychophysiological markers of anxiety and depression. Desbonnet et al. did not find a significant effect of Bifidobacteria infantis treatment on swim behaviors at day 3 or day 14 in naïve rats. However, Bifidobacteria infantis treatment attenuated pro-inflammatory immune responses and elevated tryptophan, suggesting this probiotic may possess antidepressant properties (Desbonnet et al. 2008). In their following study, they used maternal separation (MS) rats as study animal. They found MS rats reduced swim behavior and increased immobility in the forced swimming test (FST), decreased noradrenaline (NA) content in the brain, and enhanced peripheral IL-6 release and amygdala corticotrophin-releasing factor (CRF) mRNA levels compared to control rats (Desbonnet et al. 2010). Bifidobacteria infantis treatment resulted in normalization of the immune response, reversal of behavioral deficits, and restoration of basal NA concentrations in the brainstem. These changes were comparable to those observed in the citalopram group, suggesting that some probiotic benefits are similar to antidepressant effects (Desbonnet et al. 2010). Liu et al found that Lactobacillus plantarum PS128 (PS128) could ameliorate anxiety-like, but not depression-like behavior in GF mice, and significantly increased levels of both serotonin and dopamine in striatum, but not in prefrontal cortex and hippocampus (Liu WH et al. 2016). In MS model, however, PS128 significantly reduced depression-like, but not anxiety-like behaviors. PS128 significantly reduced anxiety-like, but not depression-like behavior in naïve mice. It also reduced inflammatory cytokine, IL-6, levels and increased anti-inflammatory cytokine, IL-10, level in MS mice (Liu YW et al. 2016). PS128 had more obvious effect on psychophysiology in MS mice than that in naive mice suggests that probiotic administration may have more robust effect on modification in behaviors, neurochemistry, and HPA in those suffering from higher stress.

1.3 Microbiota composition in patients with MDD Since compositions of gut microbiota alter inflammation, neurochemistry and behavior in animal studies, several human studies compared gut microbiota between patients with MDD and healthy controls. Naseribafrouei et al. compared gut microbiota compositions in 37 depressed and 18 non-depressed participants. The results showed Bacteroidales overpresentation, while Lachnospiraceae underrepresentation, in depressed participants. At low taxonomic levels, Oscillibacter and Alistipes were associated with depression (Naseribafrouei et al. 2014). Jiang et al recruited 29 active MDD (A-MDD), 17 responded MDD (R-MDD), and 30 healthy controls (HC) and compared their gut microbiota. They found that increased fecal bacterial alpha-diversity in the A-MDD vs. the HC group but not in the R-MDD vs. the HC group. Bacteroidetes, Proteobacteria, and Actinobacteria significantly increased, whereas that of Firmicutes was significantly reduced in the A-MDD and R-MDD groups compared with the HC group. MDD groups had increased levels of Enterobacteriaceae and Alistipes, but reduced levels of Faecalibacterium (Jiang et al. 2015). Not only differences in composition of gut microbiota between patients with MDD and healthy control, transplanting fecal sample from MDD patients into mice can induce depressive-like behavior in recipient mice. The composition of gut microbiota in recipient mice was similar to that of human host (Zheng et al. 2016). This study supports that gut microbiome may have a causal role in the development of depressive-like behaviors.

1.4 Effects of probiotics on mood in human studies In addition to the differences in distribution of gut microbiota between depressed subjects and non-depressed subjects, several human studies have shown that probiotics may have beneficial effect on mood. Benton et al. demonstrated that no significant difference in change of mood, measured by Profile of Mood State, between healthy participants using probiotic (containing Lactobacillus casei) milk for 3 weeks and those using milk without probiotic. However, they found that those who were in low mood at baseline became better after using probiotic-containing milk (Benton et al. 2007). Mohammadi et al. also showed that participants using probiotic yogurt (containing Lactobacillus acidophilus LA5 and Bifidobacterium lactis BB12) or probiotic capsule (containing Actobacillus casei, L. acidophilus, L. rhamnosus, L. bulgaricus, .B. breve, B. longum, S. thermophiles) for 6 weeks had significant improvement in mood, measured by GHQ and Depression Anxiety and Stress Scale, but those using conventional yogurt did not have significant change in these measurement. There were no differences in HPA axis indices, such as cortisol and ACTH, among these three groups (Mohammadi et al. 2016). Messaoudi's study, in which participants were healthy volunteer, showed that probiotics (containing L. helveticus R0052 and B. longum R0175) administration could improve participants' mood symptoms, measured by Hopkins Symptom Checklist-90 and Hospital Anxiety and Depression Scale, compared to those used placebo (Messaoudi et al. 2011). Heightened cognitive reactivity to normal, transient changes in sad mood is an established marker of vulnerability to depression and is considered an important target for interventions (Kruijt et al. 2013). Steenbergen et al. found that overall cognitive reactivity significantly reduced, which was largely accounted for by reduced rumination and aggression subscales, in participants without current mood disorders after 4 weeks multispecies probiotics administration. Because they used non-depressed participants, BDI and BAI did not significantly change after intervention (Steenbergen et al. 2015). The efficacy of probiotics in mood symptoms is not only shown in healthy volunteers, but also in clinical subjects. In patients with chronic fatigue syndrome, Rao et al. found that administration of Lactobacillus casei for 2 months could improve BAI, but not BDI, compared to placebo. At the same time fecal Lactobacillus and Bifidobacteria increased after 2-month L. casei intervention (Rao et al. 2009). Akkasheh et al. gave multi-species probiotics, containing Lactobacillus acidophilum, Lactobacillus casei, and Bifidobacterium bifidum, or placebo to patients with MDD for 8 weeks. They found that multi-species probiotics could significantly reduce BDI scores in patients with major depressive disorder (Akkasheh et al. 2016), but which strain of probiotics exerts the beneficial effect is not clear.

1.5 Childhood trauma, inflammation, and gut permeability in depression In animal study, early life stress, such as maternal separation, elevated IL-6 level and increased depressive-like behavior (increased immobility time in FST) (Liu YW et al. 2016). One recent metaanalysis showed that childhood trauma contributes to a pro-inflammatory state in adulthood (Baumeister et al. 2016). In addition, childhood trauma also increased the risk of depressive disorders (Norman et al. 2012). Metaanalyses have shown that patients with MDD have elevated serum inflammatory cytokins (Dowlati et al. 2010; Howren et al. 2009). One possible mechanism contributing to systemic inflammation in patients with MDD is bacterial translocation due to increased gut permeability (leaky gut). Figure 2 shows the relationship among childhood trauma, gut permeability, inflammation, and depression.

Every microbe possesses a microbe-associated molecular pattern (MAMP) or pathogen-associated molecular pattern (PAMP). Microbes leaking from gut activate the MAMPs and pattern-recognition receptors, such as Toll-like receptors (TLRs). Proinflammatory cytokines, such as IL-6, TNF-alpha, are released from cells after a series of signal transduction.

Probiotics may affect gut permeability. HFD-derived gut microbiota increased intestinal permeability, systemic inflammation, and brain inflammation (Bruce-Keller et al. 2015). One in-vitro study showed that Lactobacillus rhamnosus GG ameliorates gut barrier dysfunction and inflammation through inhibition of NF-kappaB (Donato et al. 2010). Administration of Lactobacillus farciminis could alleviate intestinal permeability and inflammation in partial restraint rats (Ait-Belgnaoui et al. 2012).

Evidence of increased gut permeability in patients with MDD was revealed by Maes' studies. Maes et al found that the median values for serum IgM and IgA against LPS of gram-negative enterobacteria were significantly greater in patients with MDD than in normal volunteers, suggesting that increased gut permeability results in an increased translocation of gram-negative microbiota (Maes et al. 2008; Maes et al. 2012). In patients with alcohol dependence, those with increased gut permeability, which was measured by the radioactive probe 51Cr-EDTA, had higher depression and anxiety scores than those with normal gut permeability. The composition and activity of microbiota were altered in patients with increased gut permeability (Leclercq et al. 2014).

1.6 Rationale of current study PS128 was found in traditional fermented food, fu-tsai, by our co-PI (Tsai YC) and its genome architecture has been finished and indicated its potential immunomodulatory effect. (Chao et al. 2009). Lactobacillus plantarum is generally recognized as safe (GRAS). PS128 can reduce IL-6 levels and increased IL-10 level, which accompanied with improved depression-like behavior, in MS mice (Liu YW et al. 2016). In the current study, we would like to investigate the effect of PS128 on psychophysiology, including composition of gut microbiota, gut permeability and inflammation, TLR expression, serum cytokines levels, and depressive symptoms in patients with MDD.

Metaanalyses have shown that patients with MDD have higher inflammatory cytokines (Dowlati et al. 2010; Howren et al. 2009), but only a subgroup of patients with MDD has a higher level of inflammation. For example, according to the definition of high inflammation of hs-CRP > 3 mg/l (Pearson et al. 2003), the proportion of patients with high inflammation was 24 % in patients with MDD (Rapaport et al. 2016) and 45% in patients with treatment-resistant depression (Raison et al. 2011). Baseline inflammatory level may affect treatment response. For example, patients with high baseline inflammation had better treatment response to infliximab (tumor necrosis factor antagonist) (Raison et al. 2013) and omega-3 fatty acids (Rapaport et al. 2016) compared to those with low baseline inflammation. In Raison's study, the effect of infliximab on depression symptoms was worse than placebo in those with lower inflammatory level at baseline (Raison et al. 2013). Therefore, we will recruit patients with MDD and high inflammatory level into our PS128 intervention study.

We hypothesize that PS128 can (1) change the composition of gut microbiota; (2) improve gut permeability, which is measured by fecal zonulin level, and inflammation, which is measured by fecal calprotectin level; (3) reduce the translocation of pathogen into circulation system, which will be demonstrated by reduced TLR expression, (4) reduce IL-6 and TNF-alpha, and hs-CRP, and increase IL-10 level. We hope the improvement in inflammation level can also have a beneficial effect on depressive symptoms.

Inclusion Criteria:

1. Age 20-65
2. fulfill Diagnostic and Statistical Manual of Mental Disorders fifth version (DSM-V) criteria of major depressive episode in recent 2 years
3. Psychotropics including antidepressants, antipsychotics and hypnotics have been kept unchanged for at least 3 months
4. hs-CRP > 3 mg/L
5. HAMD-17 < 14

Exclusion Criteria:

1. Comorbid with schizophrenia, bipolar disorder, or other substance use (except tobacco) disorder.
2. having active suicidal or homicidal ideation
3. known allergy to probiotics
4. comorbid with hypertension, diabetes mellitus, irritable bowel syndrome, inflammatory bowl disease, liver cirrhosis, or autoimmune diseases
5. known active bacterial, fungal, or viral infections in one month.
6. use of antibiotics, steroid, immunosuppressants, probiotics, or synbiotics in the month before collecting blood and fecal samples
7. pregnant or lactating women
8. who obviously change dietary pattern or in diet within previous two months