Tryptophan metabolism, its relation to inflammation and stress markers and association with psychological and cognitive functioning: Tasmanian Chronic Kidney Disease pilot study

This pilot study brings together data from different research areas, exploring associations between CKD progression, tryptophan metabolites, markers for inflammation and HPA activity, and psychological/cognitive functioning. There are a number of limitations of the current analysis that should be acknowledged. As this study was originally designed as a pilot to assess the feasibility of a larger cohort study, a formal power calculation was not performed. Another substantial limitation of the study design is the lack of healthy controls, dictated by the design of the Tasmanian CKD pilot study, as described in published work [39]. Results should therefore be treated as preliminary and interpreted with caution. In addition, due to the cross-sectional analysis, the direction of causality cannot be inferred. Further, it should be noted that depressive and anxiety symptoms, and cognitive functioning were assessed via self-report rather than standardized clinical interviews. While this approach may underestimate the prevalence of these disorders, all the self-report measures used in the current study have been validated in renal populations and are brief and easy to administer and interpret.

Our findings of increased tryptophan catabolites with deteriorating kidney function were mostly expected according to the literature, where available (Table 2). As we hypothesized, the high levels of the inflammation marker neopterin correlated with the tryptophan to kynurenine breakdown index. Similar correlations between KYN/TRP and neopterin were previously shown in haemodialysis patients [11] and in cases of kidney allograft rejection [15]. This was attributed to IDO-induced activation through inflammation [1, 2, 5, 7], and is in agreement with a CKD study by Schefold et al. who correlated other inflammatory markers with KYN/TRP [21]. Neopterin correlations found in the current study also included metabolites downstream in the kynurenine pathway: Quinolinic acid (as seen in [21]), 3-OH anthranilic acid, and also quinaldic acid, the redox product of kynurenic acid (showed correlation in [21]). Cortisol exhibited only non-significant trend towards negative associations with tryptophan and kynurenine, thus in this cohort we do not support the hypothesis concerning its involvement in tryptophan metabolism to kynurenine via TDO activation. This observation may, however, reflect inconsistent cortisol levels in the small cohort due to medications (i.e. ACE inhibitors), circadian variations and possible glucocorticoids resistance in some patients [44]. IDO activity may be indirectly induced by indoxyl sulfate [31], which promotes inflammation and oxidative stress [33]. The positive correlation we found between neopterin and indoxyl sulfate (R = 0.68; P < 0.0001) along with correlations between indoxyl sulfate and four kynurenines may further strengthen its link to inflammation and kynurenines production. This aspect also warrants further investigation in larger studies. With regard to the negative association we reported between indoxyl sulfate and eGFR (R = −0.71; P < 0.0001), it is debatable whether to expect a relationship with eGFR, since its clearance may be ruled by tubular secretion [45‐47] or by glomerular filtration [48]. Regardless of the clearance mechanism, its inverse relationship to eGFR was demonstrated in several studies and was related to acceleration in the progression of CKD [34]. The increased tryptophan degradation via the kynurenine pathway, as observed in CKD, was proposed to be responsible for disruptions in the 5-HT pathway due to the loss of the precursor tryptophan [5, 23‐25]. Criticisms of this hypothesis include the observation that in depressed patients, despite the inflammatory phenotype, the depletion of tryptophan was without changes in kynurenine pathway metabolites, expression of IDO or even the serotonin transporter (SERT) [49]. In our preliminary study with the low number of mostly mild depressive symptoms expressed by participants, the induced tryptophan metabolism via the kynurenine pathway did not seem to hinder metabolism via the 5-HT pathway. It is unclear whether the elevated levels of 5-OH IAA can be attributed to higher serotonin turnover or to uremic accumulation of 5-OH IAA. Nevertheless, a 9-fold increase of 5-OH IAA in CKD patients compared to healthy controls has been recorded before [50], and with our observation of association with eGFR, it may support the use of 5-OH IAA as a uraemic marker in CKD.

The second hypothesis in our work suggested association between metabolites (mainly kynurenines) and psychological/cognitive measures, possibly due to their neuroactivity. Our results revealed only a few associations between metabolites and psychological scores and subscales. As suggested in a recent review by Lopresti et al. [51], this lack of associations can result from differences between studies (type and level of psychological symptoms; comorbidities; treatment; study design; sample size). It can also originate in the involvement of additional biochemical mechanisms, not examined here, which are of high impact in CKD patients. The frequency of cognitive impairment estimated in the current study followed the expected pattern matching kidney function and treatment [52]. A low yet significant negative association was found between kynurenic acid and cognitive function, similarly to other studies [7, 24, 53], which regarded kynurenic acid as neurotoxic rather than neuroprotective (as previously hypothesized [28]). Also, we found no associations with the ratio kynurenic acid/kynurenine, termed the “neuroprotective ratio” in depression studies [28]. The neurotoxicity of kynurenic acid was partly explained by the effects of kynurenines on the nicotinic cholinergic system [54]. Kynurenic acid is a NMDA (N-Methyl-D-aspartate) receptor antagonist [5, 9, 28], unlike quinolinic acid, a NMDA receptor agonist which causes excitotoxic neurodegenerative changes [4, 7]. However, kynurenic acid has higher affinity to α-7-nicotinic acetylcholine receptors [4, 24, 54] and this can change its activity from neuroprotective to neurotoxic. The haemodialysis clearance of kynurenic acid, along with other kynurenines, is still insufficient despite advances in dialysis [12, 14, 21, 36]. The development of cognition-enhancement medication is an ongoing effort to address brain diseases, and these include pharmaceuticals that reduce the formation of kynurenic acid in the brain. Different approaches have been taken, for example by specific targeting of kynurenine aminotransferase II (KAT II, see Fig. 1) [55], which showed improved cognitive behaviour in mice lacking the expression of the enzyme [56]. Currently there is no standard treatment to eliminate kynurenines, and this emphasizes the need to reduce kynurenines formation via control of inflammation and other factors which contribute to tryptophan catabolism into kynurenines, including reduced consumption of tryptophan. An example for one approach is the treatment with selective vitamin D receptor activators (such as paricalcitol), which reduced inflammation and oxidative stress in renal patients undergoing haemodialysis [57]. Contrary to our hypothesis, mainly non-kynurenines were associated with depression and anxiety scales in our study. The self-reported complaints were mainly of low or mild symptoms, thus differing from other studies investigating tryptophan metabolism, especially in the field of psychiatry and brain diseases. Tryptophan itself did not correlate with any of the subscales, agreeing with a similar observation for a single subscale reflecting fatigue (SF-36 vitality) in haemodialysis patients [58]. 5-OH IAA along with serotonin turnover (5-OH IAA / 5-HT), were suggested before to serve as markers for evaluation of depression [59, 60]. The two markers were associated with a few subscales of depression and anxiety in our study. The multiple associations with subscales and also with total anxiety and depression scores found in patients with ESRD are of low statistical power. Nevertheless, the 5-HT pathway should not be neglected in future CKD studies in this context, also following evidence that serotonin is decreased in haemodialysis [21, 36], and data supporting the role of melatonin in fatigue symptoms of people with CKD [25]. Current guidelines recommend the use of an SSRI (selective serotonin reuptake inhibitor) as a first-line agent where treatment with antidepressants is considered in CKD patients [38]. Still, there is a lack of well-controlled trials that support or refute the efficacy and safety of antidepressant medications in patients with CKD [38, 61, 62]. This gap in medical evidence contributes to the status of clinical depression as under-recognized and an under-treated problem in this patient population. The tryptophan metabolites involving gut microflora gained our attention due to some surprising findings. Despite very high serum levels and extensive inter-metabolite correlations, indoxyl sulfate was only associated with one subscale of anxiety and not with any depressive symptoms. This, unfortunately, prevents a comparison with the intriguing report of lower serum indoxyl sulfate in depressed haemodialysis patients (compared to non-depressed patients) [63]. A novel finding in this pilot study is the role of indole-3-acetic acid (IAA) as the major metabolite related to both anxiety and depressive symptoms in CKD patients. IAA is a known uraemic molecule, and although its levels were not associated with serum creatinine or eGFR, they were elevated in some patients. IAA can be generated in the intestines from indole produced by gut microflora [30], or metabolized in tissue from tryptamine [64] and other tryptophan derivatives. IAA at uremic concentrations was previously linked to oxidative stress and contribution to development of CVD in CKD patients via activation of the Aryl Hydrocarbon Receptor [13]. The only literature we found relating IAA to psychiatry research included early studies in severely depressed patients, showing no change in urinary excretion of IAA throughout the stages of disease and recovery, despite a drop in tryptamine excretion [65] or plasma tryptophan [64]. In our work, the minimal relation between IAA and all other measured metabolites suggests the involvement of additional metabolic transformations not examined here, or subject-specific differences in its production and catabolism which obscured significant findings. Haemodialysis is rather limited in improving the clearance of indole-based uraemic toxins produced by gut microflora [33], depending on their conjugates and affinity to albumin. The intestinal absorption of indoles may be reduced by the orally-administered carbon-adsorbent AST-120. Adding this pharmaceutical to the standard therapy of CKD patients showed reduced serum levels of indoxyl sulfate [32, 66], improvement in some of the uraemia symptoms [66] and a decrease in markers for cardiovascular disease [67]. Although clinical trials showed that treatment with AST-120 resulted in more gradual disease progression [66, 68], its overall benefit is yet to be proven [68, 69]. Other approaches to reducing gut-microflora-produced uraemic molecules are consumption of a low-protein diet and restoring the balance of gut microflora by probiotics and prebiotics [30, 34, 70]. It would be interesting to examine the effects of such approaches on patients in various stages of CKD (including transplant and haemodialysis patients), and measure uraemic indole derivatives, tryptophan metabolites, inflammation markers and psychological variables.