The enduring effects of early-life stress on the microbiota-gut-brain axis are buffered by dietary supplementation with milk fat globule membrane and a prebiotic blend.

Nutritional interventions targeting the microbiota-gut-brain axis are proposed to modulate stress-induced dysfunction of physiological processes and brain development. Maternal separation (MS) in rats induces long-term alterations to behaviour, pain responses, gut microbiome and brain neurochemistry. In this study, the effects of dietary interventions (milk fat globule membrane [MFGM] and a polydextrose/galacto-oligosaccharide prebiotic blend) were evaluated. Diets were provided from postnatal day 21 to both non-separated and MS offspring. Spatial memory, visceral sensitivity and stress reactivity were assessed in adulthood. Gene transcripts associated with cognition and stress and the caecal microbiota composition were analysed. MS-induced visceral hypersensitivity was ameliorated by MFGM and to greater extent with the combination of MFGM and prebiotic blend. Furthermore, spatial learning and memory were improved by prebiotics and MFGM alone and with the combination. The prebiotic blend and the combination of the prebiotics and MFGM appeared to facilitate return to baseline with regard to HPA axis response to the restraint stress, which can be beneficial in times where coping mechanisms to stressful events are required. Interestingly, the combination of MFGM and prebiotic reduced the long-term impact of MS on a marker of myelination in the prefrontal cortex. MS affected the microbiota at family level only, while MFGM, the prebiotic blend and the combination influenced abundance at family and genus level as well as influencing beta-diversity levels. In conclusion, intervention with MFGM and prebiotic blend significantly impacted the composition of the microbiota as well as ameliorating some of the long-term effects of early-life stress.

CD4+CD25+ T Cells are Essential for Behavioral Effects of Lactobacillus rhamnosus JB-1 in Male BALB/c mice.

Over the past decade there has been increasing interest in the involvement of the microbiota-gut-brain axis in mental health. However, there are major gaps in our knowledge regarding the complex signaling systems through which gut microbes modulate the CNS. The immune system is a recognized mediator in the bidirectional communication continuously occurring between gut and brain. We previously demonstrated that Lactobacillus rhamnosus JB-1 (JB-1), a bacterial strain that has anxiolytic- and antidepressant-like effects in mice, modulates the immune system through induction of immunosuppressive T regulatory cells. Here we examined a potential causal relationship between JB-1 induced regulatory T cells and the observed effects on behaviour. We found that depletion of regulatory T cells, via treatment with monoclonal antibody against CD25, inhibited the antidepressant- and anxiolytic-like effects induced by 4-week oral administration of JB-1 in mice. Ly6Chi monocytes were found to be decreased in JB-1 fed mice with intact regulatory T cells, but not in JB-1 fed mice following depletion. Furthermore, adoptive transfer of CD4+CD25+ cells, from JB-1 treated donor mice, but not from controls, induced antidepressant- and anxiolytic-like effects in recipient mice. Ly6Chi monocytes were also significantly decreased in mice receiving CD4+CD25+ cells from JB1 fed donors. This study identifies cells within the CD4+CD25+ population, most likely regulatory T cells, as both necessary and sufficient in JB-1-induced antidepressant- and anxiolytic-like effects in mice, providing novel mechanistic insight into microbiota-gut-brain communication in addition to highlighting the potential for immunotherapy in psychiatric disorders.

Prenatal low-dose penicillin results in long-term sex-specific changes to murine behaviour, immune regulation, and gut microbiota.

Growing evidence suggests that environmental disruptors of maternal microbes may have significant detrimental consequences for the developing fetus. Antibiotic exposure during early life can have long-term effects on neurodevelopment in mice and humans. Here we explore whether exposure to low-dose penicillin during only the last week of gestation in mice has long-term effects on offspring behaviour, brain, immune function, and gut microbiota. We found that this treatment had sex-specific effects in the adult mouse offspring. Female, but not male, mice demonstrated decreased anxiety-like behaviours, while male, but not female, mice had abnormal social behaviours which correlated with altered brain expression of AVPR1A, AVPR1B, and OXTR, and decreases in the balance of splenic FOXP3+ regulatory T cells. Prenatal penicillin exposure also led to distinct microbiota compositions that clustered differently by sex. These data suggest that exposure of pregnant mice to even a low dose of penicillin through only the last week before birth is nonetheless sufficient to induce long-term sex-specific developmental changes in both male and female offspring.

Antibiotics and the nervous system: More than just the microbes?

The use of antibiotics has recently risen to prominence in neuroscience due to their potential value in studying the microbiota-gut-brain axis. In this context they have been largely employed to illustrate the many influences of the gut microbiota on brain function and behaviour. Much of this research is bolstered by the abnormal behaviour seen in germ-free animals and other well-controlled experiments. However, this literature has largely failed to consider the neuroactive potential of antibiotics themselves, independent from, or in addition to, their microbicidal effects. This is problematic, as clinical as well as experimental literature, largely neglected through the past decade, has clearly demonstrated that broad classes of antibiotics are neuroactive or neurotoxic. This is true even for some antibiotics that are widely regarded as not absorbed in the intestinal tract, and is especially concerning when considering the highly-concentrated and widely-ranging doses that have been used. In this review we will critically survey the clinical and experimental evidence that antibiotics may influence a variety of nervous system functions, from the enteric nervous system through to the brain and resultant behaviour. We will discuss substantial evidence which clearly suggests neuro-activity or -toxicity by most classes of antibiotics. We will conclude that, while evidence for the microbiota-gut-brain axis remains strong, clinical and experimental studies which employ antibiotics to probe it must consider this potential confound.

Neurobehavioural effects of Lactobacillus rhamnosus GG alone and in combination with prebiotics polydextrose and galactooligosaccharide in male rats exposed to early-life stress.

Early life is a period of significant brain development when the brain is at its most plastic and vulnerable. Stressful episodes during this window of development have long-lasting effects on the central nervous system. Rodent maternal separation (MS) is a reliable model of early-life stress and induces alterations in both physiology and behaviour. Intriguingly, the gut microbiota of MS offspring differ from that of non-separated offspring, suggesting a mechanistic role for the microbiota-gut-brain axis. Hence, we tested whether dietary factors known to affect the gut microbiota alter the neurobehavioural effects of MS. The impact of consuming diet containing prebiotics polydextrose (PDX) and galactooligosaccharide (GOS) alone or in combination with live bacteria Lactobacillus rhamnosus GG (LGG) from weaning onwards in rats subjected to early-life MS was assessed. Adult offspring were assessed for anxiety-like behaviour in the open field test, spatial memory using the Morris water maze, and reactivity to restraint stress. Brains were examined via PCR for changes in mRNA gene expression. Here, we demonstrate that diets containing a combination of PDX/GOS and LGG attenuates the effects of early-life MS on anxiety-like behaviour and hippocampal-dependent learning with changes to hippocampal mRNA expression of genes related to stress circuitry, anxiety and learning.

Growing up in a Bubble: Using Germ-Free Animals to Assess the Influence of the Gut Microbiota on Brain and Behavior.

There is a growing recognition of the importance of the commensal intestinal microbiota in the development and later function of the central nervous system. Research using germ-free mice (mice raised without any exposure to microorganisms) has provided some of the most persuasive evidence for a role of these bacteria in gut-brain signalling. Key findings show that the microbiota is necessary for normal stress responsivity, anxiety-like behaviors, sociability, and cognition. Furthermore, the microbiota maintains central nervous system homeostasis by regulating immune function and blood brain barrier integrity. Studies have also found that the gut microbiota influences neurotransmitter, synaptic, and neurotrophic signalling systems and neurogenesis. The principle advantage of the germ-free mouse model is in proof-of-principle studies and that a complete microbiota or defined consortiums of bacteria can be introduced at various developmental time points. However, a germ-free upbringing can induce permanent neurodevelopmental deficits that may deem the model unsuitable for specific scientific queries that do not involve early-life microbial deficiency. As such, alternatives and complementary strategies to the germ-free model are warranted and include antibiotic treatment to create microbiota-deficient animals at distinct time points across the lifespan. Increasing our understanding of the impact of the gut microbiota on brain and behavior has the potential to inform novel management strategies for stress-related gastrointestinal and neuropsychiatric disorders.

Reframing the Teenage Wasteland: Adolescent Microbiota-Gut-Brain Axis.

Human adolescence is arguably one of the most challenging periods of development. The young adult is exposed to a variety of stressors and environmental stimuli on a backdrop of significant physiological change and development, which is especially apparent in the brain. It is therefore unsurprising that many psychiatric disorders are first observable during this time. The human intestine is inhabited by trillions of microorganisms, and evidence from both preclinical and clinical research focusing on the established microbiota-gut-brain axis suggests that the etiology and pathophysiology of psychiatric disorders may be influenced by intestinal dysbiosis. Provocatively, many if not all of the challenges faced by the developing teen have a documented impact on these intestinal commensal microbiota. In this review, we briefly summarize what is known about the developing adolescent brain and intestinal microbiota, discuss recent research investigating the microbiota-gut-brain axis during puberty, and propose that pre- and probiotics may prove useful in both the prevention and treatment of psychiatric disorders specifically benefitting the young adult.

The TRPV1 channel in rodents is a major target for antinociceptive effect of the probiotic Lactobacillus reuteri DSM 17938.

Certain probiotic bacteria have been shown to reduce distension-dependent gut pain, but the mechanisms involved remain obscure. Live luminal Lactobacillus reuteri (DSM 17938) and its conditioned medium dose dependently reduced jejunal spinal nerve firing evoked by distension or capsaicin, and 80% of this response was blocked by a specific TRPV1 channel antagonist or in TRPV1 knockout mice. The specificity of DSM action on TRPV1 was further confirmed by its inhibition of capsaicin-induced intracellular calcium increases in dorsal root ganglion neurons. Another lactobacillus with ability to reduce gut pain did not modify this response. Prior feeding of rats with DSM inhibited the bradycardia induced by painful gastric distension. These results offer a system for the screening of new and improved candidate bacteria that may be useful as novel therapeutic adjuncts in gut pain. Certain bacteria exert visceral antinociceptive activity, but the mechanisms involved are not determined. Lactobacillus reuteri DSM 17938 was examined since it may be antinociceptive in children. Since transient receptor potential vanilloid 1 (TRPV1) channel activity may mediate nociceptive signals, we hypothesized that TRPV1 current is inhibited by DSM. We tested this by examining the effect of DSM on the firing frequency of spinal nerve fibres in murine jejunal mesenteric nerve bundles following serosal application of capsaicin. We also measured the effects of DSM on capsaicin-evoked increase in intracellular Ca(2+) or ionic current in dorsal root ganglion (DRG) neurons. Furthermore, we tested the in vivo antinociceptive effects of oral DSM on gastric distension in rats. Live DSM reduced the response of capsaicin- and distension-evoked firing of spinal nerve action potentials (238 ± 27.5% vs. 129 ± 17%). DSM also reduced the capsaicin-evoked TRPV1 ionic current in DRG neuronal primary culture from 83 ± 11% to 41 ± 8% of the initial response to capsaicin only. Another lactobacillus (Lactobacillus rhamnosus JB-1) with known visceral anti-nociceptive activity did not have these effects. DSM also inhibited capsaicin-evoked Ca(2+) increase in DRG neurons; an increase in Ca(2+) fluorescence intensity ratio of 2.36 ± 0.31 evoked by capsaicin was reduced to 1.25 ± 0.04. DSM releasable products (conditioned medium) mimicked DSM inhibition of capsaicin-evoked excitability. The TRPV1 antagonist 6-iodonordihydrocapsaicin or the use of TRPV1 knock-out mice revealed that TRPV1 channels mediate about 80% of the inhibitory effect of DSM on mesenteric nerve response to high intensity gut distension. Finally, feeding with DSM inhibited perception in rats of painful gastric distension. Our results identify a specific target channel for a probiotic with potential therapeutic properties.

Adolescence, the Microbiota-Gut-Brain Axis, and the Emergence of Psychiatric Disorders.

Second only to early life, adolescence is a period of dramatic change and growth. For the developing young adult, this occurs against a backdrop of distinct environmental challenges and stressors. A significant body of work has identified an important role for the microbiota-gut-brain (MGB) axis in the development and function of the brain. Given that the MGB axis is both highly plastic during the teenage years and vulnerable to environmental stressors, more attention needs to be drawn to its potential role in the emergence of psychiatric illnesses, many of which first manifest during adolescence. Here, we review the current literature surrounding the developing microbiome, enteric nervous system, vagus nerve, and brain during the adolescent period. We also examine preclinical and clinical research involving the MGB axis during this dynamic developmental window and argue that more research is needed to further understand the role of the MGB in the pathogenesis of brain disorders. Greater understanding of the adolescent MGB axis will open up the exciting potential for new microbial-based therapeutics for the treatment of these often-refractory psychiatric illnesses.

Squalamine reverses age-associated changes of firing patterns of myenteric sensory neurons and vagal fibres.

Vagus nerve signaling is a key component of the gut-brain axis and regulates diverse physiological processes that decline with age. Gut to brain vagus firing patterns are regulated by myenteric intrinsic primary afferent neuron (IPAN) to vagus neurotransmission. It remains unclear how IPANs or the afferent vagus age functionally. Here we identified a distinct ageing code in gut to brain neurotransmission defined by consistent differences in firing rates, burst durations, interburst and intraburst firing intervals of IPANs and the vagus, when comparing young and aged neurons. The aminosterol squalamine changed aged neurons firing patterns to a young phenotype. In contrast to young neurons, sertraline failed to increase firing rates in the aged vagus whereas squalamine was effective. These results may have implications for improved treatments involving pharmacological and electrical stimulation of the vagus for age-related mood and other disorders. For example, oral squalamine might be substituted for or added to sertraline for the aged.

Psychoactive bacteria Lactobacillus rhamnosus (JB-1) elicits rapid frequency facilitation in vagal afferents.

Mounting evidence supports the influence of the gut microbiome on the local enteric nervous system and its effects on brain chemistry and relevant behavior. Vagal afferents are involved in some of these effects. We previously showed that ingestion of the probiotic bacterium Lactobacillus rhamnosus (JB-1) caused extensive neurochemical changes in the brain and behavior that were abrogated by prior vagotomy. Because information can be transmitted to the brain via primary afferents encoded as neuronal spike trains, our goal was to record those induced by JB-1 in vagal afferents in the mesenteric nerve bundle and thus determine the nature of the signals sent to the brain. Male Swiss Webster mice jejunal segments were cannulated ex vivo, and serosal and luminal compartments were perfused separately. Bacteria were added intraluminally. We found no evidence for translocation of labeled bacteria across the epithelium during the experiment. We recorded extracellular multi- and single-unit neuronal activity with glass suction pipettes. Within minutes of application, JB-1 increased the constitutive single- and multiunit firing rate of the mesenteric nerve bundle, but Lactobacillus salivarius (a negative control) or media alone were ineffective. JB-1 significantly augmented multiunit discharge responses to an intraluminal distension pressure of 31 hPa. Prior subdiaphragmatic vagotomy abolished all of the JB-1-evoked effects. This detailed exploration of the neuronal spike firing that encodes behavioral signaling to the brain may be useful to identify effective psychoactive bacteria and thereby offer an alternative new perspective in the field of psychiatry and comorbid conditions.

Animal models of visceral pain and the role of the microbiome.

Visceral pain refers to pain arising from the internal organs and is distinctly different from the expression and mechanisms of somatic pain. Diseases and disorders with increased visceral pain are associated with significantly reduced quality of life and incur large financial costs due to medical visits and lost work productivity. In spite of the notable burden of illness associated with those disorders involving increased visceral pain, and some knowledge regarding etiology, few successful therapeutics have emerged, and thus increased attention to animal models of visceral hypersensitivity is warranted in order to elucidate new treatment opportunities. Altered microbiota-gut-brain (MGB) axis communication is central to the comorbid gastrointestinal/psychiatric diseases of which increased visceral (intestinal) sensitivity is a hallmark. This has led to a particular focus on intestinal microbiome disruption and its potential role in the etiology of heightened visceral pain. Here we provide a review of studies examining models of heightened visceral pain due to altered bidirectional communication of the MGB axis, many of which are conducted on a background of stress exposure. We discuss work in which the intestinal microbiota has either been directly manipulated (as with germ-free, antibiotic, and fecal microbial transplantation studies) or indirectly affected through early life or adult stress, inflammation, and infection. Animal models of visceral pain alterations with accompanying changes to the intestinal microbiome have the highest face and construct validity to the human condition and are the focus of the current review.

The Effect of Microbiota on Behaviour.

There is currently enormous interest in the impact of the intestinal microbiota on the development and function of the brain via activity of the microbiota-gut-brain axis. It has long been recognised that symbiotic microorganisms influence host behaviour, but in recent years evidence has accumulated that this can, in fact, be beneficial to the host. Indeed, substantial research has now demonstrated an influence of the intestinal microbiota on a wide range of mammalian behaviours. Here, we review what is currently known about the influence of intestinal microbiota on learning and memory, olfaction, social behaviours, and circadian processes. While work in animal models is compelling, further work is required to elucidate mechanisms whereby bacterial influence is occurring, as well as to determine the extent to which gut microbiota can influence similar phenotypes in humans.

Loss of vagal integrity disrupts immune components of the microbiota-gut-brain axis and inhibits the effect of Lactobacillus rhamnosus on behavior and the corticosterone stress response.

The vagus nerve is one of the major signalling components between the gut microbiota and brain. However, the exact relationship between gut-brain signaling along the vagus and the effects of gut microbes on brain function and behaviour is unclear. In particular, the relationship between the vagus nerve and immune signaling, that also appears to play a critical role in microbiota-gut-brain communication, has not been delineated. The aim of the present study was to determine the effect of subdiaphragmatic vagotomy on peripheral and central immune changes associated with the anxiolytic actions of L.rhamnosus. Male mice underwent vagotomy or sham surgery, followed by administration of L.rhamnosus for 14 days. L.rhamnosus administration following sham surgery resulted in reduced anxiety-like behaviour, and an attenuation of the hypothalamic-pituitary-adrenal axis (HPA axis), as indicated by reduced plasma corticosterone after acute restraint stress. These effects were associated with an increase in splenic T regulatory cells and a decrease in activated microglia in the hippocampus. The anxiolytic effects, HPA modulation and increase in T regulatory cells were prevented by vagotomy, whereas vagotomy alone led to a significant increase in activated microglia in the hippocampus that was not altered with L.rhamnosus treatment. Thus, both microbe induced and constitutive vagal signaling influences critical immune components of the microbiota-gut-brain axis. These findings suggest that, rather than acting as a direct neural link to the central nervous system, the role of the vagus nerve in gut-microbe to brain signalling is as an integral component of a bi-directional neuroimmunoendocrine pathway.

Membrane vesicles of Lacticaseibacillus rhamnosus JB-1 contain immunomodulatory lipoteichoic acid and are endocytosed by intestinal epithelial cells.

Intestinal bacteria have diverse and complex influence on their host. Evidence is accumulating that this may be mediated in part by bacterial extracellular membrane vesicles (MV), nanometer-sized particles important for intercellular communication. Little is known about the composition of MV from gram-positive beneficial bacteria nor how they interact with intestinal epithelial cells (IEC). Here we demonstrate that MV from Lacticaseibacillus rhamnosus JB-1 are endocytosed in a likely clathrin-dependent manner by both mouse and human IEC in vitro and by mouse IEC in vivo. We further show that JB-1 MV contain lipoteichoic acid (LTA) that activates Toll-like receptor 2 (TLR2) and induces immunoregulatory interleukin-10 expression by dendritic cells in an internalization-dependent manner. By contrast, neither LTA nor TLR2 appear to be required for JB-1 MV endocytosis by IEC. These results demonstrate a novel mechanism by which bacterial MV can influence host physiology and suggest one potential route for beneficial influence of certain bacteria and probiotics.

Identification of SSRI-evoked antidepressant sensory signals by decoding vagus nerve activity.

The vagus nerve relays mood-altering signals originating in the gut lumen to the brain. In mice, an intact vagus is required to mediate the behavioural effects of both intraluminally applied selective serotonin reuptake inhibitors and a strain of Lactobacillus with antidepressant-like activity. Similarly, the prodepressant effect of lipopolysaccharide is vagus nerve dependent. Single vagal fibres are broadly tuned to respond by excitation to both anti- and prodepressant agents, but it remains unclear how neural responses encode behaviour-specific information. Here we demonstrate using ex vivo experiments that for single vagal fibres within the mesenteric neurovascular bundle supplying the mouse small intestine, a unique neural firing pattern code is common to both chemical and bacterial vagus-dependent antidepressant luminal stimuli. This code is qualitatively and statistically discernible from that evoked by lipopolysaccharide, a non-vagus-dependent antidepressant or control non-antidepressant Lactobacillus strain and are not affected by sex status. We found that all vagus dependent antidepressants evoked a decrease in mean spike interval, increase in spike burst duration, decrease in gap duration between bursts and increase in intra-burst spike intervals. Our results offer a novel neuronal electrical perspective as one explanation for mechanisms of action of gut-derived vagal dependent antidepressants. We expect that our ex vivo individual vagal fibre recording model will improve the design and operation of new, extant electroceutical vagal stimulation devices currently used to treat major depression. Furthermore, use of this vagal antidepressant code should provide a valuable screening tool for novel potential oral antidepressant candidates in preclinical animal models.

The vagus nerve is necessary for the rapid and widespread neuronal activation in the brain following oral administration of psychoactive bacteria.

There is accumulating evidence that certain gut microbes modulate brain chemistry and have antidepressant-like behavioural effects. However, it is unclear which brain regions respond to bacteria-derived signals or how signals are transmitted to distinct regions. We investigated the role of the vagus in mediating neuronal activation following oral treatment with Lactobacillus rhamnosus (JB-1). Male Balb/c mice were orally administered a single dose of saline or a live or heat-killed preparation of a physiologically active bacterial strain, Lactobacillus rhamnosus (JB-1). 165 min later, c-Fos immunoreactivity in the brain was mapped, and mesenteric vagal afferent fibre firing was recorded. Mice also underwent sub-diaphragmatic vagotomy to investigate whether severing the vagus prevented JB-1-induced c-Fos expression. Finally, we examined the ΔFosB response following acute versus chronic bacterial treatment. While a single exposure to live and heat-killed bacteria altered vagal activity, only live treatment induced rapid neural activation in widespread but distinct brain regions, as assessed by c-Fos expression. Sub-diaphragmatic vagotomy abolished c-Fos immunoreactivity in most, but not all, previously responsive regions. Chronic, but not acute treatment induced a distinct pattern of ΔFosB expression, including in previously unresponsive brain regions. These data identify that specific brain regions respond rapidly to gut microbes via vagal-dependent and independent pathways, and demonstrate that acute versus long-term exposure is associated with differential responses in distinct brain regions.

Loss of T cells influences sex differences in stress-related gene expression.

Deficiencies in the adaptive immune system have been linked to anxiety-like behaviours and stress reactivity. Mice lacking T lymphocytes through knockout of the T cell receptor (TCR) ß and δ chains were compared to wild type C57Bl/6 mice. Central stress circuitry gene expression was assessed following repeated restraint stress. TCRß-/-δ-/- mice showed an increased baseline plasma corticosterone and exaggerated changes in stress-related gene expression after repeated restraint stress. Sexual dimorphic stress responses were observed in wild-type C57Bl/6 mice but not in TCRß-/-δ-/- mice. These data suggest that T cell-brain interactions influence sex-differences in CNS stress circuitry and stress reactivity.

Sex dependent effects of post-natal penicillin on brain, behavior and immune regulation are prevented by concurrent probiotic treatment.

There is increasing awareness of the need to consider potential long-term effects of antibiotics on the health of children. In addition to being associated with immune and metabolic diseases, there is evidence that early-life antibiotic exposure can affect neurodevelopment. Here we investigated the effect of low dose of penicillin V on mice when administered for 1 week immediately prior to weaning. We demonstrated that exposure to the antibiotic during the pre-weaning period led to long-term changes in social behaviour, but not anxiety-like traits, in male mice only. The change in behaviour of males was associated with decreased hippocampal expression of AVPR1A and AVPR1B while expression of both receptors was increased in females. Spleens of male mice also showed an increase in the proportion of activated dendritic cells and a corresponding decrease in regulatory T cells with penicillin exposure. All changes in brain, behaviour and immune cell populations, associated with penicillin exposure, were absent in mice that received L. rhamnosus JB-1 supplementation concurrent with the antibiotic. Our study indicates that post-natal exposure to a clinically relevant dose of antibiotic has long-term, sex dependent effects on the CNS and may have implications for the development of neuropsychiatric disorders. Importantly, we also provide further evidence that probiotic based strategies may be of use in counteracting detrimental effects of early-life antibiotics on neurodevelopment.

Publisher Correction: Increased persistence of avoidance behaviour and social deficits with L.rhamnosus JB-1 or selective serotonin reuptake inhibitor treatment following social defeat.

An amendment to this paper has been published and can be accessed via a link at the top of the paper.