The association of physical and sexual trauma to future development of functional disorders such as IBS has long been recognized.4, 5 Although the reasons for this relationship remain unclear, heightened awareness to visceral and somatic symptoms is speculated to cause reporting of minor symptoms.5, 6 A similar process may account for the higher prevalence of psychological and psychiatric disorders observed in IBS patients: depression, somatization disorder, generalized anxiety disorder, panic and phobic disorders and coping difficulties are described in the literature.7, 8 These observations coupled with advances in brain imaging technology, functional MRI and proton emission tomography (PET), suggest regional brain activation at sites linked to affect and attention as precursors of IBS symptoms. Studies have consistently shown differences in regional cortical and subcortical activation between IBS subjects and healthy controls in response to visceral stimulation. For example, IBS patients showed predominant activation within the insula, prefrontal cortex and thalamus9, 10 and enhanced activity within the cingulate cortex10–12 following visceral stimulation. Overall, the enhanced activity occurs in cortical areas involved in affect and attention and subcortical areas associated with arousal and autonomic responses. Anticipation of non-noxious or noxious stimuli resulted in similar patterns of cortical activation, supporting the concept of an emotional component to perception.11, 13 Habituation, which can be described as decreased perception to recurrent aversive visceral stimuli without IBS symptoms intensity changes, implicates decreased vigilance or altered afferent input as modulators of the perceptual process.13
Hyper-vigilance with over-reporting of symptoms may account for the commonly alluded notion of visceral hypersensitivity as a part of IBS pathogenesis, i.e. the reporting of pain in response to stimuli, which are innocuous to healthy volunteers.14 Dorn et al. investigated the contribution of over-reporting in IBS using sensory decision theory analysis (SDT), a process which determines the physiological and psychological components of pain thresholds.15 In the study, statistical methods were used to calculate a Discriminant Index as a measure of neurosensory sensitivity (physiological component) and a Report Criterion (psychological component), a measure of subjects’ overall labelling of symptoms as weak or intense independent of the stimulus intensity, which is susceptible to modulation by cognitive and psychological factors. Calculations were carried out on initial measurements of pain and urge thresholds and subsequently analysed for the physiological and psychological components of each thresholds. As anticipated, IBS patients reported lower pain and urge thresholds compared with healthy controls. SDT analysis revealed a greater tendency to report pain, but similar neurosensitivity compared with non-IBS subjects. Both pain and urge thresholds correlated with the report criterion, but not with neurosensitivity index implying that IBS subjects’ heightened colonic sensitivity was caused by psychological component rather than physiological factors. However, the limiting factor of the study remains the reliance on verbal symptom descriptions without a direct neurophysiological measure. An objective instead of a descriptive marker of central cerebral activity during visceral stimulation is needed to support the above findings, to differentiate increased afferent central input from enhanced perception as a result of hyper-vigilance or over-reporting.
With functional MRI as a tool, Lawal et al. showed greater cortical activity in IBS compared with controls in response to subliminal (sub-conscious) stimulation.16 His observations are in contradiction to the findings of Dorn et al. showing a contributory part of neurosensitivity in the form of enhanced activity with central neural networks independent of cognitive function because of subliminal nature of the applied stimulus. Other methods, such as cortical evoked potentials (CEP), which demonstrate detailed and dynamic analysis of the central cerebral activity in response to stimuli, may shed further light on the central pathways in IBS.
Cortical evoked potentials in response to visceral stimuli exhibit three identifiable peak deflections captured by scalp electrodes: the first positive peak P1 followed by a negative peak N1 and a second positive peak P2 (Figure 1). Analysis of waveform characteristics in healthy volunteers as a response to stimulus intensity showed a positive relationship between stimulus intensity and the amplitude of P1–N1 and N1–P2. Furthermore, the latency of P1 decreases as the stimulus intensity increases suggesting more active recruitment of sensory afferent fibres (Figure 2).17 In noncardiac chest pain, a distinct functional disorder characterized by oesophageal hypersensitivity, CEPs successfully differentiated distinctive phenotypes according to pain thresholds and CEP waveforms changes.18 Subjects with reduced pain threshold showed either a normal latency with enhanced wave amplitude suggestive of true visceral hypersensitivity or increased latency and reduced wave amplitude suggestive of over-symptoms reporting without enhanced central transmission. Whilst studies applying this method of measuring cortical activity in IBS subjects are underway, alternative methods with functional MRI have shown greater sub-conscious stimuli-induced cortical activity in IBS compared with controls support a contributory role of central neural network hyper-reactivity independent of cognitive function.16
Figure 1. Oesophageal cortical evoked potential (CEP). This figure demonstrates a rectal CEP acquired in one healthy female subject in response to electrical stimulation. There are three main components to the CEP: the P1, N1 and P2 peaks. There is a strong correlation between values of amplitude and latency of these components and subjective sensory ratings for the experimental stimulus. (Adapted from Hobson et al. 2000, Am J Physiol Gastrointest Liver Physiol).18
Download figure to PowerPoint
Figure 2. Rectal cortical evoked potentials (CEPs) with increasing stimulus intensity. This figure demonstrates CEPs acquired from a healthy volunteer with increasing intensity of electrical stimulation. Five stimulation intensities were used in this study ranging from sensory threshold through to pain. As the stimulus intensity ascends from the sensory threshold towards the pain threshold, there is a corresponding increase in the amplitude and decrease in latency of the CEP. The early part of the rectal CEP response (<250 ms, represented by the shaded area) represents stimulus-specific processing within a host of cortical regions providing information about the sensitivity of the visceral afferent pathway. Beyond 250 ms we observe stimulus related, endogenous brain activity (labelled LR – Late response), which is similar to the P300 reported in many other cognitive studies. It is of note that the late response only really develops when the stimulus is quite definitely perceived and engages higher order brain processes (Adapted from Harris et al.17).
Download figure to PowerPoint
Despite access to sophisticated imaging technologies, the understanding of pain mechanisms in functional gut disorders (FGD) is still at an infancy stage. Dissection of pathways linking higher cortical function with emotive/attentional/perceptual factors to the final expression of symptoms is the key to further our understanding. To date, studies have relied detecting differences on fMRI and PET activation using different experimental paradigms and analysis techniques.19 Although the activated areas were comparable between different studies, emotional and cognitive aspects of visceral sensation were not considered in interpretation of results. Furthermore, most studies have used rectosigmoid distension as the visceral stimulus to investigate central processing of pain, but in humans, it is unlikely that symptoms are solely because of distension. Instead, sensory afferent fibres activated by chemoreceptors, osmoreceptors, metaboreceptors and nocioceptors collectively converge on the dorsal horn. The final expression of cortical sensation from all these receptors is determined by modulatory effects of central inhibitory and facilitatory neurones on the afferent activity in the dorsal horn. Translation of finding of one stimulus to a disease state may be inapplicable to other stimuli. Moreover, valid conclusions pertaining to therapeutic interventions rely on the ability to sample neuro-sensory activity on multiple occasions in the same subject to reduce intra and inter individual variability of the final outcome measure. Two studies have examined test re-test variability using fMRI. In the first, Coen et al. used the same volume oesophageal balloon distension to induce cortical activity in healthy volunteers on three separate study days.20 These data showed that whilst activation was fairly consistent in primary/secondary sensory areas and the thalamus, activation in the cingulate cortex diminished between days1 and 3. This habituation of activity in the cingulate cortex was associated with a reduction in the subjective pain rating of the stimulus and was most likely related to the reduction of procedure-related anxiety over time. A similar longitudinal study in IBS patients using PET revealed that whilst the cortical activation pattern remained stable over time, activity in key limbic regions diminished and this was also associated with normalization of sensory reports of rectal distension thresholds.13 These studies provide a cautionary reminder that extraneous factors such as endogenous pain modulation and state anxiety can significantly alter brain activity and should be considered as variables when statistically modelling brain imaging data. Although it may be easier to undertake such repetitive examinations in animal studies, such an approach may be counter-intuitive as sub-primate animal only utilizes subcortical mechanisms to drive sensory and autonomic reflexes.
Irritable bowel syndrome probably represents a heterogenous group of patient with similar symptoms. Collating data from brain imaging studies to generate significant results may be masking significant observations evident only by individualized analysis of CEP, fMRI and PET results to identify patterns to generate sub-groups. Sub-group identification will be the future challenge for IBS and although brain imaging will play a fundamental role to develop this concept, confounders such as attention, expectation and hypervigilance will have an inevitable impact on outcomes. Now that direct visualization of activated brain regions in response to stimulus is possible, the approach to better understanding of the brain–gut axis has now shifted to identifying networks between these regions. Mayer et al. describe a homeostatic afferent processing network that maintains the body’s physiological status.21 More recently, a visceral pain network has been traced using fMRI and Diffusion Tensor Imaging technique, which allow fibre tracking to show activation of neural fibres between cortical brain regions.22 Re-focusing imaging efforts to answer specific mechanistic questions may reinvigorate this area of research.
Several studies suggested that primary afferent neurones (PAN) are hypersensitive to non-noxious stimuli in IBS (visceral hypersensitivity) with lower sensory thresholds to rectosigmoid balloon distension.23 Using rapid phasic colorectal distension to elicit an aversive sensation response (discomfort), Mertz et al. showed lower threshold for discomfort in IBS compared with healthy controls, whilst stool thresholds and thresholds to slow ramp distension were similar.24 Studies carried out in our unit suggest modulation of visceral pain by stress: visceral pain thresholds during physical or psychological stress were lower in IBS patients compared with healthy volunteers.25 Unfortunately, the scientific merit of recto-sigmoid distension tests is mitigated by confounding factors such as anticipatory responses, sensitization to repeated distensions,26 level anxiety or stress,25 distensions protocols,27 age,28 gender29 and the intensity of the IBS symptoms24 may all affect the reproducibility and hence the outcome measure.
At a cellular level, peripheral sensitization arises when an inflammatory or injurious event releases mediators that activate and sensitize nociceptive afferent nerves.30 The stimulus is transmitted via spinal afferent neurones (ANs) and to a lesser extent vagal afferents, which primarily exert modulator function. Within the dorsal horn, spinal afferents integrate with Lamina I neurons, which also receive descending (inhibitory and facilitatory) fibres from the brainstem and hypothalamus.31 This dynamic process permits modulation of sensation by emotions and cognition and provides a mechanism whereby anxiety and stress alter symptom perception.
There is mounting evidence for low-grade inflammatory process as a contributor to gut dysfunction. Research in this field has rapidly gathered momentum over the past 2 years. The initial studies showed qualitative and quantitative cellular changes in IBS compared to healthy controls. Chadwick et al. described increased intestinal CD3+ lymphocyte counts in colonic lamina propria of IBS patients compared to controls.32 Colonic biopsies collected from patients with post-infectious IBS demonstrated an increased number of T cells and macrophages.33, 34 Other researchers showed an increase in the number of mast cells within the colonic mucosa and muscularis propria.35, 36
Studies measuring inflammatory cytokine levels support the role of immune system activation in IBS pathogenesis. Elevated blood concentrations of IL-6, as well as IL-1β and TNFα, were observed in post-infectious and diarrhoea-predominant IBS.36, 37 A strong correlation between symptom severity and levels of the pro-inflammatory cytokines TNFα and IL-1β implies a direct effect of cytokines on visceral sensation. O’Mahoney et al. adopted a different approach to investigate the immune system in IBS.38 In a comparative controlled study whereby peripheral cytokine levels were measured before and after probiotic therapy, the authors demonstrated an increase in IL-10 to IL-12 ratio and an accompanying reduction in symptoms score after probiotic therapy.
A genetic susceptibility has been proposed as a putative mechanism for the altered cytokine levels. When genotypes for high and low IL-10 production were studied, IBS patients were less likely than healthy controls to have the genotype for high IL-10 production with a loss of its anti-inflammatory properties.39 Instead, an increase in mucosal rectal pro-inflammatory IL-1β mRNA expression was noted in post-infectious IBS patients compared to nongastroenteritis IBS controls.40
Liebreits studied the relationship between cytokine levels, symptoms and psychiatric morbidity. Baseline levels of pro-inflammatory cytokines TNFα, IL-1β and IL-6 and LPS-stimulated release from peripheral blood mononuclear cells (PBMC) were linked to the clinical presentation of IBS, implying inflammatory cytokines in symptom manifestation of IBS.37 Furthermore, a significant direct relationship between anxiety and LPS-induced TNFα production was reported. Recent reports show less pro-inflammatory IFNγ and anti-inflammatory IL-10 secretion from PBMC collected form IBS patients compared with controls.41
Ohman et al. recently demonstrated that IBS patients exhibit an increased frequency of integrin β7 T cells in peripheral blood. β7 integrin mediates homing of these cells to the intestinal mucosa.42 The gut homing pattern of lymphocytes in IBS suggests that after encountering their specific antigen in a gut-associated lymph node, T cells home to the intestinal mucosa to undertake their effector function.
The relationship between cytokine and symptom generation is critical to substantiate the pathogenetic role of inflammation in IBS. Initial suggestions of a direct link were reported by Barbara’s group, which showed the presence of mast cells in close proximity to enteric nervous system.35 The investigators presented stronger evidence by correlating intensity of pain with the number of mast cells. Preliminary reports for animal studies support a direct effect of cytokines and other mediators like histamine, on mucosal spinal ANs. When supernatants of PBMC taken from patients with IBS were applied to rodent afferent nerves, a 60% increase in afferent activity was noted compared to supernatant from healthy controls.43 Although it remains unclear whether the cytokine changes represent a primary or secondary process and how the magnitude of the alterations is reflected in visceral sensation, there does seem to be a direct effect on afferent activity to account for symptoms.
The relationship between symptoms and cytokines can also be inferred by studying other disease models characterized by florid inflammation, e.g. ulcerative colitis. Active ulcerative colitis has been shown to be associated with rectal hypersensitivity in a few studies.44, 45 When only subjects with mild inflammation were studied, rectal perception was unexpectedly found to be attenuated compared to IBS.46 Preserved activation of anti-nociceptive mechanisms with mild inflammation may account for the unexpected finding and perhaps the nature of the inflammatory infiltrate may regulate pain perception.
The number of gut bacteria has been estimated to be 100 trillion organisms.47 The experimental tools required for in depth analysis of the intestinal microbiota are only now becoming available and the status of the commensal flora in IBS remains to be defined. Nonetheless, evidence suggests that intestinal bacteria play a significant role in inducing IBS. The original observations come from studies on post-infectious IBS whereby a subgroup of patients developed IBS following a bout of gastroenteritis.48, 49 Subsequently, other studies reported the onset of functional symptoms following antibiotic treatment for either extra-intestinal infections50 or gastroenteritis.51 Some investigators take these observations further by proposing bacterial overgrowth as a cause for IBS. The justification rests on abnormal lactulose breath tests results noted in several IBS patients supported by symptom improvement after antibiotic therapy.52, 53 Lactulose breath tests are notoriously prone to false positive results especially in the context of dysmotility associated with IBS and so far no specific bacterial species has yet been linked to the disease. Whilst this may reflect the difficulties of specific species identification because of microbiotal diversity and nonculturable organisms, these limitations have been overcome by using molecular techniques. Sequencing bacterial 16S rRNA genes from clones obtained by PCR amplification of faecal DNA is feasible and has been adopted to study microbiota in healthy subjects.54 However, the PCR amplification process creates a bias towards dominant low Guanine plus cytosine bacterial species.55 Fractionating the DNA preparations according to the sample G + C percentage (%G + C) allows less abundant species as well as sequences with high G + C contents to be amplified by PCR. Fractionation according to the %G + C content thus enriches the diversity of sequences obtained for cloning and sequencing. Application of this technique to IBS patients and healthy controls showed a difference in profiles between these two groups and within IBS subsets depending on the predominate symptom.56 The bacterial 16S rRNA genes from the selected differing GC profiles were subsequently cloned and partially sequenced to obtain more detailed information of the composition of microbial populations. Differences between the clone libraries were noted for several bacterial genera, some of which were verified by quantitative PCR. The alterations involved several bacterial genera with Lactobacillus sequences absent from IBS and Collinsella sequences greatly reduced in IBS. In future, sophisticated metagenomic analyses will allow greater gene sequencing from clones to reveal patterns in IBS.
An alternative simpler approach adopted to study the impact of microbiota on IBS symptoms has been to manipulate gut bacteria by probiotic ingestion. Clinical IBS probiotic trials have so far revealed discordant clinical responses. O’Mahoney et al. reported alleviation of multiple IBS symptoms with Bifidobacterium infantis, but not with Lactobacillus salivarus.38 Two separate placebo-controlled studies reported improvement in bloating, flatulence and colonic transit in response to the probiotic cocktail VSL#3, which comprises eight different bacterial strains including bifidobacteria and lactobacilli.38, 57, 58 Similarly, two other studies with Lactobacillus species showed an improvement in pain and flatulence.59, 60 In contrast, two other trials with Lactobacillus did not report benefit; one of these was a cross over design, a study design which has limitations for IBS studies.36, 61 The largest study with Bifidobacteria showed that it may be particularly effective in alleviating IBS symptoms.62 An increase in the peripheral anti-inflammatory (IL-10) and pro-inflammatory (IL-12) cytokine ratio was associated with probiotic therapy and symptom improvement.38 The translation of a luminal stimulus into an immunological response implies communication of between microbiota and the immune system. The ability of gut dendritic cells (DCs) to recognize and respond in a highly specific manner to different bacteria has led to the concept that they may be one of the targets of probiotic bacterial therapies63 as shown in other models of gut diseases.
In patients suffering from IBD, altered cytokine production from colonic activated DC suggests involvement of DC in the pathogenesis of this disease.64 Previous studies also showed that blood and colonic mucosal DC from healthy volunteers upregulated IL-10 production and downregulated IL-12 production in response to Bifidobacteria65 present in the VSL#3 cocktail. Dietary supplementation with a bifidogenic fructoligosaccharide modulated the activity of colonic DC in Crohn’s disease patients in vivo.66 Other researchers have reported similar anti-inflammatory effects of probiotic bacteria on peripheral DC collected from patients with active colitis.67 It seems that commensal bacteria may help maintain an anti-inflammatory balance between IL-12 and IL-10 and that disruption of the equilibrium, for example, by reducing luminal Bifidobacteria, induces inflammation. A schematic overview of the interactions among microbiota, immune system, sensory afferent and higher cortical areas is shown in Figure 3. Although intestinal DC in IBS has not been studied yet, there is sufficient evidence to justify further exploration of the interplay between microbiota and DC in IBS. The justification for this research approach comes from animal models described in the next section.
Figure 3. This figure demonstrates the putative relationship between microbiota, immune cells, sensory afferents and the central nervous system: (1) Dendritic cell (DC) sample gut bacteria directly, or via M cells and epithelial cells, and respond to them using pathogen recognition receptors. Following migration to draining lymphoid tissue, gut DC activate and shape the adaptive immune response via the production of cytokines (2). The nature of the response induced (e.g. pro-inflammatory vs. regulatory) can be influence by the nature of bacteria sampled. DC also target lymphocytes back to the intestine by inducing expression of homing receptors (3). Here, reactivation by antigen-bearing antigen presenting cell including DC drives the local release of pro- and anti-inflammatory cytokines with balance determined by a combination of local factors and the properties of DC during priming in the lymph nodes. Inflammatory cytokine production recruits and activates additional immune cells including APC, mast cells and PMNs. It can also increase intestinal permeability leading to increased access of bacterial products to immune cells and further rounds of recruitment and activation (4). Depending on the balance of bacterial stimulation, the net result is the production of inflammatory mediators and activation of nociceptors on spinal afferent neurons, which relay the stimulus to the brain via the spinal cord (5). Stimulus perception can be modulated at the level of the spinal cord via Lamina I neurons (stippled area) in the dorsal horn, which receive modulator fibres from brainstem of cortex (6). AN, afferent neurone; PMNs, polymorphonuclear cells.
Download figure to PowerPoint
Animal models linking pathogenetic pathways in IBS
A comprehensive exploration of interaction between enteric bacteria, immune system and the neural system is fundamental to formulate a unifying hypothesis incorporating the findings and observations of IBS studies to date. Recent experimental studies on animal models have shed further light on how these processes may be linked. For instance, mice exposed to antibiotic therapy exhibited heightened sensitivity to colorectal distension with subsequent normalization of pain threshold following probiotic therapy. This implies modulation of visceral sensation by gut bacteria.68 The investigators proceeded to show increased inflammatory activity with antibiotic therapy, which resolved after probiotic ingestion, further strengthening the evidence for relationship of the inflammatory process to neural function.
A nociceptive effect of probiotics was the focus of a series of elegant experiments by a separate group of researchers.69 They reported an increased in epithelial cell surface expression of both opoid and cannabinoid receptors following Lactobacillus acidophilus exposure. The in vitro experiments were complimented by in vivo studies examining the response of mice to pain: lactobacilli exposure increased pain threshold to colorectal distension to an extent comparable to 1 mg of morphine.
In contrast, other studies examined a putative direct relationship between immune cell types and sensory perception by manipulating lymphocytes. Mice lacking T and B lymphocytes were found to exhibit lower pain threshold than mice with normal lymphocytes.70 Pain sensation normalized upon restoration the CD4+ T-cell population. The anti-nociceptive effect of lymphocytes was postulated to be related to endorphin receptors subsequently found on T lymphocytes surface.
Finally, an inflammatory neuro-regulatory role has been described in animal studies.71, 72 The cholinergic anti-inflammatory pathway whereby vagus nerve stimulation inhibited pro-inflammatory cytokines release, including TNFα and IL-1, suggests a restraining effect on immune activation.73 This anti-inflammatory activity represents the efferent limb of the inflammatory reflex. The afferent limb, activated by inflammatory products and mucosal-microbiota interactions,74 is thought to transmit information to the central nervous system via the vagus, which activates several systems including the hypothalamic pituitary adrenal axis (HPA) and autonomic nervous system leading to, amongst other things, acetylcholine release from vagal nerve endings. Animal experiments with vagal nerve stimulators were effective in suppressing cytokine mediated inflammation, corroborating the anti-inflammatory pathway.72, 75 Extrapolating these finding to IBS will undoubtedly advance our understanding of visceral pain in this condition.
Psychoneuroimmunological translational research
The animal studies described in the previous section point towards a link among visceral sensation, immune activation and microbiota, where manipulating enteric bacteria was associated with alteration both in mucosal cytokines and in visceral perception. Interaction between microbiota and antigen presenting cells (APCs) has been extensively studied in other inflammatory disease models like IBD as previously described, but not in IBS.63 A recent study on maternal separation rat models incorporates the psychological or stress-induced link into the overall understanding of the disease pathogenesis.76 The authors demonstrated increased intestinal permeability by maternal separation induced stress. Elevated cortisol levels caused by HPA activation and corticotropin-releasing factor (CRF) release were noted in parallel with enhanced permeability. Probiotic therapy administered prior to maternal separation mitigated the intestinal permeability change with an associated reduction in cortisol levels. The authors speculated on the mechanisms of probiotic-induced changes: the initial stress-induced enhanced permeability and the HPA activation permitted mucosal immune cell activation by gut microbiota to elicit an inflammatory response. Inflammatory cytokines may facilitate the entry of unfavourable bacteria, perpetuating the inflammatory process and further increasing the permeability. A change in microbiota with probiotic therapy may alter the interaction with mucosal immune cells so that only nonharmful bacteria interact with mucosal APCs, minimizing the inflammatory effect. Such a plausible hypothesis amalgamates the different elements within psychoneuroimmunology of IBS towards a cohesive explanation of how psychiatric or stressful life events predispose to IBS, how psychological therapies may reduce HPA activation and permeability, the role of probiotics and the visceral sensation aggravated by inflammatory mediators.