Brain Kynurenine Pathway and Functional Outcome of Rats Resuscitated From Cardiac Arrest.

Background Brain injury and neurological deficit are consequences of cardiac arrest (CA), leading to high morbidity and mortality. Peripheral activation of the kynurenine pathway (KP), the main catabolic route of tryptophan metabolized at first into kynurenine, predicts poor neurological outcome in patients resuscitated after out-of-hospital CA. Here, we investigated KP activation in hippocampus and plasma of rats resuscitated from CA, evaluating the effect of KP modulation in preventing CA-induced neurological deficit. Methods and Results Early KP activation was first demonstrated in 28 rats subjected to electrically induced CA followed by cardiopulmonary resuscitation. Hippocampal levels of the neuroactive metabolites kynurenine, 3-hydroxy-anthranilic acid, and kynurenic acid were higher 2 hours after CA, as in plasma. Further, 36 rats were randomized to receive the inhibitor of the first step of KP, 1-methyl-DL-tryptophan, or vehicle, before CA. No differences were observed in hemodynamics and myocardial function. The CA-induced KP activation, sustained up to 96 hours in hippocampus (and plasma) of vehicle-treated rats, was counteracted by the inhibitor as indicated by lower hippocampal (and plasmatic) kynurenine/tryptophan ratio and kynurenine levels. 1-Methyl-DL-tryptophan reduced the CA-induced neurological deficits, with a significant correlation between the neurological score and the individual kynurenine levels, as well as the kynurenine/tryptophan ratio, in plasma and hippocampus. Conclusions These data demonstrate the CA-induced lasting activation of the first step of the KP in hippocampus, showing that this activation was involved in the evolving neurological deficit. The degree of peripheral activation of KP may predict neurological function after CA.

1‐methyl‐DL‐tryptophan

3‐hydroxyanthranilic acid

5‐hydroxytryptamine

arginase 1

cardiac arrest

indoleamine 2,3‐dioxygenase

interleukin 1 beta

N‐methyl‐d‐aspartate

kynurenine

pathway

kynurenic acid

neurological deficit score

post–return to spontaneous circulation

repeated measures

return to spontaneous circulation

Clinical Perspective

What Is New?

Kynurenine pathway is an emerging key component of postcardiac arrest syndrome, affecting brain injury and survival after cardiac arrest.

This study demonstrates early, prolonged activation of the kynurenine pathway in the blood and brain after successful resuscitation and its involvement in the evolving neurological deficits.

What Are the Clinical Implications?

The study results point to new neuroprognostic biomarkers and therapeutic targets for cardiac arrest–induced brain injury.

Mortality is high after cardiac arrest (CA) as a consequence of the “post‐CA syndrome,” a complex pathophysiological process following the return of spontaneous circulation (ROSC). It is characterized by myocardial dysfunction with circulatory shock, systemic inflammation with activation of the clotting system, and brain injury.

Cardiovascular failure accounts for most deaths in the first 3 days after ROSC, while brain injury accounts for later deaths in two thirds of patients after out‐of‐hospital CA and in approximately a quarter of patients after in‐hospital CA. As many as 30% of CA survivors experience permanent brain damage, and as many as two thirds of resuscitated patients are discharged from hospital with variable degrees of neurological dysfunction including seizures, myoclonus, memory impairment, cognitive dysfunction, fatigue, emotional behavior, posttraumatic stress symptoms, and difficulties in managing daily activities. , , , ,

Brain injury after CA is caused by the brain’s vulnerability to the ischemic/reperfusion insult, which leads to a complex cascade of cellular and molecular events, eg, activation of microglia and astrocytes, infiltration of circulating macrophages, excitotoxicity, cytokines release, disrupted calcium homeostasis, free radical formation, and pathological protease cascades, , exacerbating the ischemic/reperfusion injury. To predict the functional/neurological outcome a multimodal approach, which also includes the analysis of circulating biomarkers, is needed due to the complexity and interplay events occurring in the post‐CA phase. , , Thus, a better understanding of the underlying pathological molecular mechanisms is required to identify new, specific prognostic biomarkers and develop targeted therapeutic approaches.

The kynurenine pathway (KP) is emerging as one of the potential key components of brain injury after CA. KP is the major pathway of catabolism of the essential aminoacid tryptophan, which is converted first into kynurenine (through the intermediate L‐formylkynurenine) by the rate‐limiting enzyme indoleamine 2,3‐dioxygenase (IDO) , (the pathway is schematically shown in Figure S1). IDO is widely expressed in a variety of human tissues, as well as in circulating macrophages and dendritic cells, and is stimulated by proinflammatory cytokines, lipopolysaccharides, and free radicals. KP activation has been described in various clinical conditions, including cardiovascular diseases, cancer, diabetes, infections, depression, schizophrenia, neurodegenerative disorders, sepsis, , and after cardiac bypass and thoracic surgery. We recently found early and prolonged activation of peripheral KP in patients resuscitated from out‐of‐hospital CA, related to the severity of post–CA shock, early death in the intensive care unit, and 12‐month poor long‐term outcome. Activation of KP has also been implicated in situations causing neurological injury in patients in the intensive care unit and in the adverse prognosis in patients with stroke and trauma.

Several downstream metabolites of kynurenine have important activities in the brain. Some have neurotoxic/excitotoxic properties such as 3‐hydroxyanthranilic acid (3‐HAA) that auto‐oxidize, leading to cerebral oxidative stress, and quinolinic acid, which acts as a potent agonist of N‐methyl‐d‐aspartate (NMDA) receptors. The main neuroprotective metabolite is kynurenic acid (KYNA), which is an antagonist at NMDA receptors, , AMPA receptors, alpha 7 nicotinic acetylcholine receptors, and the aryl hydrocarbon receptor agonist.

The present study investigated the activation of KP in CA‐induced brain injury, measuring tryptophan for the first time and the KP metabolites kynurenine, KYNA, and 3‐HAA directly in the hippocampus and in plasma of rats resuscitated from CA at different times post‐ROSC. As proof of concept, we then checked the effect of the competitive IDO inhibitor, 1‐methyl‐DL‐tryptophan (1‐DL‐MTRP), in preventing KP‐induced neurological deficit in a more severe experimental model of CA.

Methods

Data Availability

The data that support the findings of this study are available from the corresponding author upon reasonable request.

Animal Care

Procedures involving animals were conducted at the Istituto di Ricerche Farmacologiche Mario Negri IRCCS, which adheres to the principles set out in the following laws, regulations, and policies governing the care and use of laboratory animals: Italian Governing Law (D.lgs 26/2014; Authorisation n.19/2008‐A issued March 6, 2008, by the Ministry of Health); Mario Negri Institutional Regulations and Policies providing internal authorization for persons conducting animal experiments (Quality Management System Certificate – UNI EN ISO 9001:2015 – Reg. N° 6121); the National Institutes of Health Guide for the Care and Use of Laboratory Animals (2011 edition), and EU directives and guidelines (EEC Council Directive 2010/63/UE). The study also followed the ARRIVE criteria. A minimum of 9 rats per timing of measurements had been planned; the number of animals was reduced to a minimum necessary in order to be able to appreciate a significant difference of 50%, with a power of 0.8 and an error of probability of 0.05 (free software G*Power) taking into account the variability, in terms of KP metabolites levels, previously observed.

Aim No. 1: Systemic and Central KP Activation After CA and Cardiopulmonary Resuscitation

The experimental design is illustrated in Figure 1A. The aim of this study was to investigate the systemic and central (ie, hippocampus) activation of the KP after CA followed by cardiopulmonary resuscitation (CA/CPR). In parallel, evaluation of the hemodynamic response was examined. A full description of the methods for animal preparation, CA/CPR procedure, and measurements are given in Data S1.

Figure 1

Experimental designs.

Experimental designs adopted for (A) aim no. 1 and (B) aim no. 2. *Blood (0.6 mL) drawn for measurement of plasma 1‐methyl‐DL‐tryptophan, kynurenine pathway metabolites, and plasma high‐sensitivity cardiac troponin T. BL=baseline (before cardiac arrest [pre‐CA]). AP indicates arterial pressure; CPR, cardiopulmonary resuscitation; Echo, transthoracic echocardiography; NDS, neurological deficit score; RAP, right atrial pressure; ROSC, return of spontaneous circulation; and VF, ventricular fibrillation.

Animal Preparation

Male Sprague‐Dawley ex‐breeder rats (weight, 481±6 g; n=37) (Envigo) were used for the study. Thirty‐two rats were anesthetized and instrumented for hemodynamic measurements and induction of CA, as previously reported. CA/CPR was performed in 28 rats, while CA was not induced in 4 rats, which were included in the sham group. Five naive (nonoperated) rats were used as additional controls. Sham and naive rats were used for hippocampal analysis, as described in the Results section.

Experimental Procedures

We used an established model of electrically induced CA/CPR. Briefly, animals were subjected to 7 minutes of untreated CA and 5 minutes of CPR, including mechanical chest compression, ventilation with oxygen, and defibrillation. Blood samples were serially collected from the femoral artery cannula in 3K‐EDTA tubes 15 minutes before CA (pre‐CA), and 10 minutes and 2 hours after ROSC. Rats were then euthanized with an intraperitoneal injection of pentobarbital sodium (150 mg/kg), and the brain was carefully removed, dissected to isolate hippocampal region, and immediately frozen in liquid N2 before storage at −80 °C.

Measurements

Hemodynamics were recorded as previously described. Briefly, ECG, aortic pressure, and right atrial pressure were continuously monitored for up to 2 hours after ROSC on a personal computer–based data acquisition system supported by CODAS hardware and software (DataQ). Coronary perfusion pressure was calculated in the same time range as the difference between time‐coincident diastolic aortic and right atrial pressures.

Aim No. 2: Effect of IDO Inhibitor on CA‐Induced KP Activation and Neurological Deficit

The experimental design is illustrated in Figure 1B. The aim of this study was to investigate whether pretreatment with 1‐DL‐MTRP, an inhibitor of the enzyme IDO‐1, interferes with KP activation after CA/CPR and prevents neurological injury. A full description of the methods for animal preparation, CA/CPR procedure, and measurements are given in Data S1.

Animal Preparation and Pharmacological Treatment

Male Sprague‐Dawley ex‐breeder rats (weight, 480±5 g; n=42) (Envigo) were used. Six rats were used as naive controls and 36 were anesthetized and instrumented for hemodynamic measurements and induction of CA. Before the induction of anesthesia, the 36 rats were randomized using no transparent envelopes and assigned into 2 experimental groups:

Treatment group (1‐DL‐MTRP, n=17), receiving 2 equal doses of the IDO inhibitor 1‐DL‐MTRP by gavage, 16 and 2 hours before CA; 1‐D‐MTRP (Sigma Aldrich Italia) and 1‐L‐MTRP (Sigma Aldrich Italia) were dissolved in 1% carboxymethylcellulose and given orally by gavage to a final dose of 800 mg/kg of racemate.

Control group (vehicle, n=19), receiving an equal volume of 1% carboxymethylcellulose in deionized water by gavage, 16 and 2 hours before CA.

Animals were subjected to 8 minutes of untreated CA and 8 minutes of CPR, including mechanical chest compression, ventilation with oxygen, and defibrillation. This is a longer and more severe CA experimental model than that used for aim no. 1, with the intent to exacerbate KP activation. The CA/CPR procedure was performed by investigators blinded to the 2 experimental groups. Four hours after ROSC, all of the catheters and the endotracheal tubes were removed. The animals were returned to their cages and were observed up to 96 hours after resuscitation. Animals were given ampicillin (50 mg/kg intramuscular) and buprenorphine (0.16 mg/kg) to prevent infection and pain. Blood samples were serially collected from the femoral artery cannula in 3K‐EDTA tubes 15 minutes pre‐CA, and 2 and 96 hours post‐ROSC (PR 2 hours and PR 96 hours). At 96 hours, rats were euthanized with an intraperitoneal injection of pentobarbital sodium (150 mg/kg), and the brain was removed, dissected to isolate hippocampus and cortex, and immediately frozen in liquid N2 before storage at −80 °C.

ECG, aortic pressure, and right atrial pressure were continuously monitored for up to 4 hours post‐ROSC, as described for aim no. 1. Myocardial function was assessed at 4 and 96 hours post‐ROSC by transthoracic monodimensional, bidimensional, pulse, color, and tissue Doppler echocardiography (Aloka SSD‐5500). Left ventricular ejection fraction and the septum mitral inflow ratio (ie, the ratio of the early transmitral pulsed Doppler inflow velocity to peak early diastolic mitral annulus velocity by tissue Doppler) were measured and calculated to evaluate LV systolic and diastolic function, respectively. For echocardiographic analysis at 96 hours, rats were anesthetized with intraperitoneal thiopental 50 mg/kg and, if necessary, 10 mg/kg recall doses. Plasma high‐sensitivity cardiac troponin T (hs‐cTnT) concentration was assayed at pre‐CA and 4 and 96 hours post‐ROSC with an electrochemioluminiscence assay (ECLIA, Elecsys 2010 analyzer; Roche Diagnostics) following manufacture instructions.

Neurological Assessment

Neurological deficit score (NDS) was used to assess neurologic recovery 24, 48, 72, and 96 hours post‐ROSC. In brief, the test, ranging from 0 (normal) to 500 (brain death), rated level of consciousness, respiration, motor and sensory functions, and overall behavior. The score was decided by collaborators blinded to the study groups.

Inflammatory Gene Expression

Total RNA was extracted form brain cortex with a commercial kit (Pure Link RNA Mini Kit, Ambion) and processed for real‐time reverse transcription polymerase chain reaction analysis of gene expression levels for interleukin 1 beta (IL1β) and Arginase 1 (Arg1). Glyceraldehyde 3‐phosphate dehydrogenase (Gapdh) was used as the reference gene, and primer sets were designed to span exon junctions (https://www.ncbi.nlm.nih.gov/tools/primer‐blast/).

KP Metabolites and Serotonin Measurements in Plasma and Hippocampus by High‐Performance Liquid Chromatography‐Tandem Mass Spectrometry

Blood samples were centrifuged at 2000g for 15 minutes at 4 °C and plasma was stored at −80 °C until analysis. On the day of analysis, hippocampi were thawed, accurately weighed, and homogenized in 5 volumes of H2O/acetonitrile (1:2, v/v). Plasma and hippocampal levels of 1‐DL‐MTRP (aim no. 2), tryptophan (aims no. 1 and no. 2) and its metabolites, kynurenine (aims no. 1 and no. 2), KYNA (aim no. 1), 3‐HAA (aim no. 1), and 5‐hydroxytryptamine (5‐HT; aim no. 2 in hippocampus) were measured using high‐performance liquid chromatography coupled with tandem mass spectrometry. Briefly, deuterated internal standards were added to plasma and hippocampal homogenate; plasma samples were then mixed with cold methanol and incubated at −20 °C for protein precipitation before centrifugation; an appropriate volume of hippocampal homogenate was centrifuged directly. After centrifugation, supernatant was dried under N2 and the residue was resuspended and injected into the high‐performance liquid chromatography system. Separation was performed following a gradient elution and mass spectrometric analysis was done with a triple quadrupole mass spectrometer in positive ion mode and multiple reaction monitoring mode, measuring the fragmentation products of the deprotonated pseudomolecular ions. A full description of the high‐performance liquid chromatography coupled with tandem mass spectrometry methods used for aims no. 1 and no. 2 is given in Data S1.

Statistical Analysis

The GraphPad Prism program (GraphPad Software) was used for data processing and statistical analysis. A 1‐sample Kolmogorov–Smirnov Z normality test was used to inspect normal distribution of the data. Continuous variables were analyzed by 1‐way repeated‐measures (RM) ANOVA followed by Tukey multicomparison test for normally distributed data; non‐normally distributed data were analyzed with Kruskal‐Wallis test (for non‐RM) or Friedman test (for RM) followed by Dunn multicomparison test. T test or Mann‐Whitney U test (2‐tailed P value, 95% CI), respectively, was used for comparison between 2 groups of normally and non‐normally distributed data. Two‐way ANOVA followed by Sidak multiple comparisons test was used: variables non‐normally distributed were corrected by logarithmic (natural log) transformation before 2‐way ANOVA with Sidak multiple comparisons test. Ordinary (non‐RM) or RM 2‐way ANOVA were used when appropriate, as reported in the legends of the figures and tables.

When the dependent variable was categorical, Fisher exact test was used. Spearman rank correlation coefficient was used to analyze correlations between variables. P<0.05 was considered significant.

Results

Aim no. 1: Survival and Resuscitation Outcome

CA was induced in 28 rats, 23 of which (82%) were successfully resuscitated. Table S1 reports hemodynamic variables and plasma levels of hs‐cTnT.

Each rat developed marked post‐ROSC hemodynamic instability during the 2‐hour observation, as expected. Heart rate was significantly reduced 10 minutes post‐ROSC compared with baseline (P<0.0001 versus pre‐CA) then returned to pre‐CA values (Table S1). Systolic atrial pressure, mean arterial pressure, diastolic arterial pressure, and coronary perfusion pressure remained low for 2 hours post‐ROSC (P<0.0001, Table S1).

Circulating levels of hs‐cTnT reflected the development of postresuscitation myocardial injury. Median plasma levels of hs‐cTnT were significantly higher at 10 minutes (P<0.0001) and 2 hours (P=0.0002) post‐ROSC compared with pre‐CA levels (Table S1).

Aim no. 1: CA/CPR Alters KP Metabolites Levels in Plasma and Hippocampus

The effect of CA on plasma tryptophan and kynurenines is reported in Table 1. Ten‐minute post‐ROSC tryptophan levels were 12% lower (P=0.0059) and kynurenine levels were 36% higher (P=0.0001). The early activation of the tryptophan→kynurenine step of the KP is highlighted by the significant increases of the kynurenine/tryptophan ratio (P<0.0001). Regarding downstream KP metabolites, KYNA levels showed a 5‐fold increment (P<0.0001), while 3‐HAA levels decreased slightly but not significantly. Two hours post‐ROSC, kynurenine levels were still higher than pre‐CA (P=0.0012), suggesting that the tryptophan→kynurenine step remained activated (1.3‐fold increase of kynurenine/tryptophan ratio, P=0.0007 versus pre‐CA). At the same time, KYNA levels were still high (P=0.0035 versus pre‐CA), although with a tendency to decrease from the 10‐minute post‐ROSC values (P=0.0550).

Table 1

Plasma and hippocampal KP activation after CA/CPR

Plasma		Hippocampus
Metabolite	Median [quartile 1–quartile 3]	Metabolite	Median [quartile 1–quartile 3]
Tryptophan, µg/mL		Tryptophan, µg/g
Pre‐CA	17.00 [13.95–20.63]	Controls	5.05 [4.80–5.20]
PR 10 min	14.65 [13.00–18.05]*
PR 2 h	15.59 [13.73–16.87]	PR 2 h	8.09 [7.60–8.98]||
Kynurenine, µg/mL		Kynurenine, µg/g
Pre‐CA	0.561 [0.374–0.690]	Controls	0.052 [0.042–0.076]
PR 10 min	0.704 [0.541–0.957]†
PR 2 h	0.700 [0.493–0.886]†	PR 2 h	0.134 [0.096–0.169]||
Kynurenine/tryptophan ratio		Kynurenine/tryptophan ratio
Pre‐CA	0.030 [0.023–0.040]
PR 10 min	0.051 [0.039–0.061]†	Controls	0.011 [0.009–0.015]
PR 2 h	0.044 [0.036–0.065]†	PR 2 h	0.015 [0.013–0.020]$
KYNA, ng/mL		KYNA, ng/g
Pre‐CA	10.04 [7.92–4.77]	Controls	2.43 [2.08–3.25]
PR 10 min	57.03 [39.16–116.2]†
PR 2 h	17.80 [11.18–40.43]†	PR 2 h	6.95 [5.75–10.83]||
3‐HAA, ng/mL		3‐HAA, ng/g
Pre‐CA	2.137 [1.733–2.526]	Controls	0.222 [0.177–0.239]
PR 10 min	1.780 [0.962–2.227]
PR 2 h	2.762 [1.667–3.532]‡	PR 2 h	0.301 [0.276–0.545]||

Plasma concentrations of tryptophan, kynurenine, the kynurenine/tryptophan ratio, kynurenic acid (KYNA), and 3‐hydroxyanthranilic acid (3‐HAA) were measured in operated rats before cardiac arrest (pre‐CA) and 10 minutes and 2 hours post–return to spontaneous circulation (PR 10 minutes and PR 2 hours) (n=23 rats). For hippocampal analysis, tryptophan and kynurenine pathway (KP) metabolites were measured in naive rats (n=5) and in sham‐operated rats (n=4) and 2 hours PR (2 hours, n=23). Since there were no statistical differences between naive and sham rats (absolute levels and statistical analysis are reported in Table S2), these data were pooled as a control group (n=9) to evaluate the effect of cardiac arrest (CA). Effect of CA: absolute plasma levels were analysed with Friedman test for repeated measures, followed by Dunn multiple comparisons post hoc test, *P<0.01 and † P<0.001 vs pre‐CA; ‡ P<0.01 vs PR 10 minutes; absolute hippocampal levels were analyzed with Mann‐Whitney U test for nonrepeated measures: $ P<0.05 and || P<0.001 vs controls. Exact P values for each comparison are reported in the text. CPR indicates cardiopulmonary resuscitation.

KP values were preliminarily analyzed in the hippocampus of naive and sham rats (Table S2). Hippocampal levels of tryptophan and kynurenines metabolites were not significantly altered by surgical procedures alone (anesthesia and instrumentation as described in the Methods section); naive and sham rats were therefore pooled and considered as controls for subsequent analyses. Two hours post‐ROSC, hippocampal tryptophan levels had risen 60% (P<0.0001), in contrast with the decrease in plasma. The effect of CA on hippocampal kynurenine levels was even greater (+159%, P<0.0001), with an increase of the kynurenine/tryptophan ratio (+36%, P=0.0145). Median KYNA and 3‐HAA levels also increased by 186% (P=0.0001) and 36% (P<0.0001), respectively.

Aim no. 2: Effect of IDO Inhibition

The 2 oral doses of 800 mg/kg of 1‐DL‐MTRP given 16 and 2 hours before CA resulted in similar plasma concentrations at 15 minutes pre‐CA and 2 hours post‐ROSC, with an average of 34 µg/mL (range, 20–40 µg/mL), which declined slightly to 21 µg/mL on average 96 hours post‐ROSC (Figure S2A). At this last time, hippocampal concentrations of the drug were 11.4 µg/g on average (Figure S2B), with a tissue‐to‐plasma ratio of ≈0.54.

Effects on Survival, Hemodynamic, and Myocardial Function

Treatment with 1‐DL‐MTRP had no effect on body weight, heart rate, hemodynamics, or circulating biomarkers in the 2 groups pre‐CA (Table 2). There were no differences in coronary perfusion pressure during CPR (Table 2), reflecting adequate standard chest compression quality in both groups. The duration of CPR and the number of defibrillations delivered to achieve ROSC were lower in animals treated with 1‐DL‐MTRP than those given vehicle; however, these differences were not statistically significant (Table 2). Pretreatment with 1‐DL‐MTRP had no effect on the number of successfully resuscitated animals (1‐DL‐MTRP 88%, vehicle 84%) and that survived up to 96 hours after CA (1‐DL‐MTRP 66%, vehicle 62%).

Table 2

CPR Outcomes, Hemodynamics, and Myocardial Functions in CA/CPR Rats Treated With 1‐DL‐MTRP

Outcomes	Vehicle (n=19)	1‐DL‐MTRP (n=17)
Body weight, g	480±24	481±25
Resuscitation, no./total	16/19	15/17
Time to ROSC, s	481 [478–482]	481 [370–482]
Shocks to ROSC, no.	1	1
4‐h survival, no./resuscitated	16/16	15/15
96‐h survival, no./resuscitated	10/16	10/15
CPP, mm Hg
Pre‐CA	111±18	109±13
PR 10 min	44±17	57±21
PR 1 h	76±12	84±9
PR 2 h	71±14	81±12
PR 3 h	71±18	75±15
PR 4 h	71±17	73±20
HR, beats per min
Pre‐CA	368±34	375±34
PR 10 min	194±57	237±54*
PR 1 h	335±37	335±12
PR 2 h	338±36	360±20
PR 3 h	350±26	358±23
PR 4 h	352±33	351±51
SAP, mm Hg
Pre‐CA	152±9	151±13
PR 10 min	88±20	103±17*
PR 1 h	111±12	117±9
PR 2 h	110±15	118±12
PR 3 h	107±14	111±18
PR 4 h	106±11	115±17
DAP, mm Hg
Pre‐CA	118±10	114±12
PR 10 min	44±17	57±21
PR 1 h	77±13	84±9
PR 2 h	72±15	81±12
PR 3 h	70±18	74±15
PR 4 h	70±16	78±15
MAP, mm Hg
Pre‐CA	127±12	127±13
PR 10 min	59±19	76±20*
PR 1 h	95±30	96±9
PR 2 h	86±15	96±13
PR 3 h	83±16	89±16
PR 4 h	83±16	94±17
LVEF, %
PR 4 h	33.6±3.6	36.8±3.2
PR 96 h	75.1±3.2	75.3±3.2
E/e’ ratio
PR 4 h	14.9±1.0	16.4±1.4
PR 96 h	15.9±2.6	16.2±1.3
hs‐cTnT, pg/mL
Pre‐CA	100 [71–190]	98 [66–111]
PR 4 h	5995 [4942–9604]	4965 [2978–5242]
PR 96 h	64 [51–82]	64 [36–278]

Body weight (normal), resuscitation outcomes (non‐normal), and survival rate (Fisher test). Hemodynamic parameters (normal) and myocardial functions were recorded up to 4 and 96 hours post–return of spontaneous circulation (PR), respectively. All data are shown as mean±SD when normally distributed or median [quartile 1–quartile 3] when non‐normally distributed and analyzed by ordinary (for left ventricular ejection fraction [LVEF], septum mitral inflow (E/e’), and plasma high‐sensitivity cardiac troponin T [hs‐cTnT]) or repeated measures (for coronary perfusion pressure [CPP], heart rate [HR], systolic arterial pressure [SAP], diastolic arterial pressure [DAP], and mean arterial pressure [MAP]) 2‐way ANOVA followed by Sidak multicomparison post hoc test, * P<0.05 vs time‐matched vehicle‐treated rats (effect of 1‐methyl‐DL‐tryptophan [1‐DL‐MTRP]). Time and shock to return of spontaneous circulation (ROSC) were analyzed by the Mann‐Whitney U test. CA indicates cardiac arrest; and CPR, cardiopulmonary resuscitation.

After resuscitation, animals treated with 1‐DL‐MTRP had higher systolic and mean diastolic pressures and coronary perfusion pressure compared with vehicle; systolic arterial pressure and mean arterial pressure were significantly higher 10 minutes post‐ROSC (PR 10 minutes) and coronary perfusion pressure was markedly higher 2 hours after resuscitation in the 1‐DL‐MTRP–treated group compared with the vehicle‐treated group. At 4 hours post‐ROSC (PR 4 hours) in both study groups, left ventricular ejection fraction was severely compromised (normal reference value, 87.0%±1.7%) and septum mitral inflow ratio was slightly high (normal reference value, 12.0±0.9). At 96 hours post‐ROSC (PR 96 hours), left ventricular ejection fraction returned to normal in both groups and the septum mitral inflow ratio did not change. There were no significant differences in plasma levels of hs‐cTnT in the vehicle‐ and 1‐DL‐MTRP–treated groups (Table 2).

Effect on CA/CPR‐Induced KP Activation

Fifteen minutes pre‐CA, tryptophan levels in plasma of 1‐DL‐MTRP–treated rats were 27% higher than in vehicle‐treated rats (P=0.0263), while kynurenine levels were 24% lower (P=0.0243). The kynurenine/tryptophan ratio therefore decreased 40% (P=0.0056) (Table S3). These data indicate inhibition of peripheral IDO activity by 1‐DL‐MTRP before CA.

In vehicle‐treated rats, we confirmed early CA‐induced peripheral KP activation (Figure 2), ie, a significant decrease in tryptophan plasma levels (22%, P=0.0444) and a significant increase in kynurenine plasma levels (33%, P=0.0461) 2 hours post‐ROSC (PR 2 hours). At 96 hours post‐ROSC, tryptophan levels returned to pre‐CA levels (Figure 2A), but kynurenine was still increased (+56%, P=0.0022) (Figure 2B). The kynurenine/tryptophan ratio was higher at 2 hours (+68%, P=0.0128) and 96 hours post‐ROSC (+80%, P=0.0106) (Figure 2C).

Figure 2

Indoleamine 2,3‐dioxygenase inhibition counteracts cardiac arrest/cardiopulmonary resuscitation (CA/CPR)–induced kynurenine pathway (KP) activation in plasma.

Plasma levels of (A) tryptophan, (B) kynurenine, and (C) kynurenine/tryptophan ratio 15 minutes before cardiac arrest (pre‐CA; n=13–15), 2 hours post–return of spontaneous circulation (PR; PR 2 hours [n=13–15]), and 96 hours PR (n=9) in vehicle‐ and 1‐methyl‐DL‐tryptophan (1‐DL‐MTRP)–treated rats. Box plots show median, quartile range, and minumim–maximum range. Statistical analysis was performed on absolute values (pre‐CA, PR 2 hours, and PR 96 hours) by ordinary 2‐way ANOVA, followed by Sidak multiple comparisons post hoc test. *P<0.05 and **P<0.01 vs corresponding pre‐CA levels (effect of CA). + P<0.05 and ++ P<0.01 vs time‐matched vehicle‐treated rats (effect of 1‐DL‐MTRP pretreatment).

In 1‐DL‐MTRP–treated rats, we found much lower effects of CA. In particular, following CA, plasma kynurenine levels and kynurenine/tryptophan ratio showed a slight, nonsignificant increase in comparison to pre‐CA conditions, and, most important, they were markedly lower than in vehicle‐treated rats at the same time points (Figure 2).

Hippocampal kynurenine levels of vehicle‐treated rats were 109% higher than in naive rats (P=0.0064) 96 hours post‐ROSC, resulting in a higher kynurenine/tryptophan ratio (+57%, P=0.0204) and indicating sustained activation of central KP (Figure ). Kynurenine levels in 1‐DL‐MTRP–treated rats were significantly lower than in vehicle‐treated rats (P=0.0409) and comparable to those in naive rats, resulting in a lower kynurenine/tryptophan ratio (P=0.0249) than for the vehicle‐treated group) (Figure 3).

Figure 3

Indoleamine 2,3‐dioxygenase inhibition counteracts CA/CPR–induced kynurenine pathway activation in hippocampus.

Concentrations of tryptophan, kynurenine, and the kynurenine/tryptophan ratio measured in hippocampus 96 hours post–return of spontaneous circulation (96 hours) in rats treated with vehicle (n=9) or 2 doses of 800 mg/kg 1‐methyl‐DL‐tryptophan (1‐DL‐MTRP) given by gavage 16 and 2 hours before CA (n=9). Since this experiment included 2 independent sessions, each 1 including 3 naive rats, for each session we normalized tryptophan, kynurenine, and the kynurenine/tryptophan ratio in vehicle‐ and 1‐DL‐MTRP–treated rats to the means for the corresponding naive rats (absolute levels are reported in Table S4). Box plots show median, quartile range, and minimum–maximum range. Data were analyzed by Kruskal‐Wallis test followed by Dunn multicomparison test. *P<0.05 and **P<0.01 vs naive. + P<0.05 vs vehicle time‐matched. Exact P values are reported in the text.

An important finding is that 96 hours post‐ROSC, hippocampal levels of tryptophan, kynurenine, and the kynurenine/tryptophan ratio positively correlated with the levels of the same analytes in plasma (Figure 4), both in vehicle‐ and 1‐DL‐MTRP–treated rats.

Figure 4

Hippocampal kynurenine pathway metabolites levels positively correlate with those in plasma.

Correlations between hippocampal levels of (A) tryptophan, (B) kynurenine, and (C) the kynurenine/tryptophan ratio 96 hours post–return of spontaneous circulation in vehicle‐ (teal blue circles, n=9) and 1‐methyl‐DL‐tryptophan–treated rats (red circles, n=9). Correlations were calculated by Spearman method: rank correlation coefficient r [95% CI] and P value (2‐tailed) of the correlations are indicated inside the figures. P<0.05 was considered significant.

In the same rats we also found that hippocampal 5‐HT concentrations in vehicle‐ and 1‐DL‐MTRP–treated rats were superimposable to those measured in naive rats (Figure S3 and Table S4).

Effect on Neurological Functions and Central Inflammatory Response

Figure 5A reports the NDS measured 24, 48, 72, and 96 hours post‐ROSC in rats pretreated with vehicle or 1‐DL‐MTRP. NDS values were lower in resuscitated rats pretreated with the IDO inhibitor than vehicle‐treated rats, at all time points.

Figure 5

Time course of the neurological deficit score (NDS) after CA/CPR and correlation with hippocampal and plasmatic levels of kynurenine pathway metabolites.

A, NDS was calculated 24, 48, 72, and 96 hours post–return of spontaneous circulation (PR; PR 24 hours, PR 48 hours, PR 72 hours, and PR 96 hours) in rats treated by gavage 16 and 2 hours before CA with vehicle (n=9) or 1‐methyl‐DL‐tryptophan (1‐DL‐MTRP; 800 mg/kg per dose [n=9]). Box plots show median, quartile range, and minimum–maximum range. Data were analyzed by 2‐way repeated measures ANOVA, followed by Sidak multiple comparisons post hoc test. + P<0.05 vs time‐matched vehicle‐treated rats. Panels B‐E show correlations between NDS and hippocampal (B and C) and plasmatic (D and E) levels of kynurenine and the kynurenine/tryptophan ratio 96 hours PR in vehicle‐ (teal blue circles) and 1‐DL‐MTRP–treated rats (red circles). Correlations were calculated by Spearman method: rank correlation coefficient r [95% CI] and P value (2‐tailed) of the correlation are indicated inside the figures. P<0.05 was considered significant. Correlations analysis between tryptophan, in plasma and hippocampus, and NDS are reported in Figure S4.

We found a significant positive correlation between the severity of the neurological deficit (96 hours post‐ROSC in vehicle‐ and 1‐DL‐MTRP–treated rats) and kynurenine hippocampal levels (Figure 5B), as well as the kynurenine/tryptophan ratio (Figure 5C). Positive correlations were also found between NDS and plasma kynurenine levels (Figure 5D) and the kynurenine/tryptophan ratio (Figure 5E).

Further, hippocampal levels of 5‐HT did not correlate with the NDS measured 96 hours post‐ROSC (data not shown).

We then examined inflammatory gene expression by real‐time reverse transcription polymerase chain reaction in the brain cortex 96 hours post‐ROSC (Figure S5). No statistically significant differences have been observed, although there was a tendency in the upregulation of IL1β and Arg1 after CA/CPR. Pretreatment with 1‐DL‐MTRP slightly reduced CA‐induced IL1β upregulation, with a tendency towards an increase in Arg1 gene expression.

Discussion

In this study we show that: (1) CA‐induced activation of the KP, previously observed in plasma from animal models and humans and confirmed here in rats, is accompanied by parallel increases of KP metabolites in the hippocampus; (2) the increase of kynurenine, the metabolite resulting from IDO‐mediated tryptophan conversion, is long‐lasting and still present 96 hours after CA in plasma and hippocampus; (3) CA induces a clear neurological deficit in rats, which is counteracted by inhibition of IDO, suggesting the involvement of KP, which is further supported by (4) the positive correlation between neurological deficit and hippocampal kynurenine levels; and (5) the positive correlation between neurological deficit and plasma kynurenine levels also suggests that peripheral activation of KP is a marker of the functional status of the brain.

In the first experiment, we find in the plasma of a large number of rats, early (as soon as 10 minutes) and prolonged (up to 96 hours) activation of the first step of the KP (Table 1 and Figure 2) is most likely caused by IDO activation induced by the peripheral inflammatory and immune response after CA/CPR. , , , , , We also document for the first time a rapid (10 minutes) and marked (5‐fold) increase of the downstream KP metabolite KYNA, but not of 3‐HAA, with a partial return toward pre‐CA levels at 2 hours. The increase of circulating KYNA has been associated with anti‐inflammatory , and antioxidative activities. These results confirm and extend previous data in small and large animals, as well as humans, after CA/CPR. , , A similar CA‐induced alteration of peripheral KP metabolites between rats and humans has also recently been confirmed with a different rat model of CA, thus supporting the use of rats as a suitable model for studying human CA.

We also analyzed, for the first time, the effect of CA on KP metabolites in the rat hippocampus, the brain region mainly affected by ischemic/reperfusion injury after successful resuscitation after CA, in humans , and animal models. , , At variance with the results in plasma, hippocampal tryptophan concentrations were higher 2 hours post‐ROSC (+69%), returning to pre‐CA levels after 96 hours (Table 1 and Figure 3). Since tryptophan is an essential aminoacid, the rise in hippocampal levels after CA/CPR may be associated with an increase in blood‐to‐brain passage. Circulating tryptophan crosses the blood‐brain‐barrier in the free form through the large neutral amino acid transporter L‐type amino acid transporter 1. Therefore, the higher hippocampal tryptophan levels after CA/CPR might be attributable to CA‐induced regional blood‐brain‐barrier breakdown , , that facilitate the influx of the brain‐permeable metabolites , and also of circulating macrophage infiltration, , , thus increasing the influx of peripheral tryptophan. CA/CPR also markedly raised hippocampal kynurenine levels by 167% and 109% at 2 and 96 hours post‐ROSC, respectively (Table 1 and Figure 3). Rodent studies suggest that ≈60% of brain kynurenine comes from peripheral sources through the same mechanisms as tryptophan , ; therefore, the increase in hippocampal kynurenine absolute levels may be caused by higher influx into the CNS as for tryptophan. However, considering that the CA‐induced increase of hippocampal tryptophan dropped from 2 and 96 hours post‐ROSC (69% and 10%, respectively), whereas in the same period the kynurenine/tryptophan ratio rose (36% and 57%), it follows that central IDO is most likely activated after CA/CPR.

The activation of the first step of the KP in the hippocampus also caused perturbation of the neuroactive KP downstream metabolites, and hippocampal KYNA (neuroprotective) and 3‐HAA (neurotoxic) concentrations also rose 2 hours post‐ROSC. The changes in KP metabolites levels might be ascribed to fluctuations in the activity of downstream KP enzymes caused by the proliferation/activation of resident microglia and astrocytes. , , , , , , After CA/CPR, the neuroinflammatory response following brain hypoxia and reperfusion is also orchestrated by activated glial cells. , , , , , The changes in the neuroglial population might lead to unbalance in the downstream KP metabolites since the enzymatic steps kynurenine→KYNA and kynurenine→3‐HAA (Figure S1) are physically segregated into 2 distinct cellular branches. Finally metabolized to quinolinic acid, 3‐HAA is synthetized in the microglia, while kynurenine is converted to KYNA in astrocytes. , , ,

The activation of the first step of the KP and the fluctuations in the levels of KYNA and 3‐HAA suggest a perturbation of the whole pathway, with changes in the hippocampal levels of the other neuroactive metabolites. Higher levels of 3‐HAA are generally associated with higher levels of quinolinic acid, which in addition to its activity as a potent agonist of the NMDA receptor, also raises the levels of reactive oxygen and nitrogen species, stimulates lipid peroxidation, and causes mitochondrial dysfunctions. , It is also involved in deregulation of the phosphorylation state of several cellular proteins. ,

We then confirm the involvement of KP activation on the central consequences of CA, assessed on the basis of the NDS at different times after ROSC. Our approach was to prevent the first step of KP, pretreating rats with 1‐DL‐MTRP, a well‐known IDO inhibitor. , , We used the dose of 800 mg/kg, since previous studies in animal models (rats and dogs) showed no further increase in the plasma concentration of 1‐DL‐MTRP with higher oral doses, because of saturable absorption and the absence of toxicity. 1‐DL‐MTRP plasma levels after 2 oral doses 16 hours and 2 hours before CA confirmed the attainment of steady‐state and slow elimination from blood. , Plasma and brain levels were, respectively, 21 µg/mL and 11.4 µg/g 96 hours post‐ROSC, indicating good brain penetration of the inhibitor, in line with previous data. , These drug levels, corresponding to concentrations in the high micromolar range (52–96 µmol/L), are compatible with inhibition of the enzyme, since in vitro studies showed a Ki of 34 µmol/L. ,

Pretreatment with 1‐DL‐MTRP therefore significantly lowered plasma levels of kynurenine and of the kynurenine/tryptophan ratio, indicating successful inhibition of the first step of KP. In these conditions, we observed significant improvement of the neurological function after CA, in agreement with our hypothesis of involvement of the KP in the central events leading to brain injury after CA/CPR. Preclinical and clinical studies in stroke have in fact shown that interfering with the KP may have beneficial effects in terms of neuroprotection and reduction of neuroinflammation, lessening patient morbidity and mortality. Our data indicate that pretreatment with 1‐DL‐MTRP reduced kynurenine levels in plasma and hippocampus 96 hours post‐ROSC, and, consequently, the kynurenine/tryptophan ratio.

To investigate the effect of IDO inhibition on the hypoxic ischemic brain environment, we analyzed gene expression of factors associated with proinflammatory states and anti‐inflammatory states, IL‐1β, and Arg1, respectively. , , The results are not conclusive but are suggestive for indicating an effect of 1‐DL‐MTRP in counteracting the CA/CPR‐induced increase of IL‐1β mRNA and in increasing Arg1 mRNA.

We did not find effects of CA/CPR on hippocampal 5‐HT levels, a possible consequence of the shift toward the KP induced by IDO activation (Figure S3), nor were changes observed after treatment with 1‐DL‐MTRP. However, these measurements were performed 96 hours after ROSC, and it cannot be excluded that 5‐HT alterations might occur at earlier time points. Moreover, the rapid turnover of 5‐HT to form its downstream metabolites (eg, 5‐hydroxyindoleacetic acid) could mask the effect of CA/CPR or IDO inhibition on central 5‐HT levels.

Interestingly, despite interindividual variability in kynurenine levels and consequently in the kynurenine/tryptophan ratio after CA/CPR, we found significant correlations between individual kynurenine absolute levels, as well as kynurenine/tryptophan ratio and CA‐induced neurological scores. These correlations were observed in hippocampal and plasma levels, suggesting that peripheral activation of KP may be a marker of functional outcome, consistent with previous findings in out‐of‐hospital CA patients and in patients after stroke, where peripheral activation of the KP correlated with stroke severity and cerebral infarct volume. , Moreover, the correlation between plasma and hippocampal levels of kynurenine and kynurenine/tryptophan ratio also suggest that peripheral KP may reflect the situation in the brain.

In this study, we did not measure the hippocampal levels of KYNA and 3‐HAA in the animals treated with 1‐DL‐MTRP, since the main aim was to prove the involvement of the first step of the KP pathway in CA‐induced neurological impairment. It might be envisaged that the inhibition of the KP pathway should also result in a decrease of downstream metabolites, including the neuroprotective KYNA, raising doubts on the role of this metabolite in the beneficial effects of 1‐DL‐MTRP. It might be considered, however, that many sets of data , , actually show that 1‐MT treatment induces a counterintuitive increase of plasmatic KYNA, while producing, as expected, an increase in tryptophan and a decrease in kynurenine and in kynurenine/tryptophan ratio, with no change of 5‐HT levels. These data suggest alternative mechanisms of KYNA production, possibly mediated by other tryptophan‐degrading enzymes or nonenzymatically. This additional mechanism of action of 1‐DL‐MTRP, and its involvement for the reduction of CA‐induced neurological deficits, requires further investigations.

We recognize some limitations in our study. We investigated the changes in KP activity in the periphery and in the brain by measuring absolute levels of tryptophan, kynurenine, KYNA, and 3‐HAA, and we suggest activation of KP based on the ratio between metabolites. However, IDO activation is not the only determinant regulating the peripheral kynurenine/tryptophan ratio. Other factors include: (1) the activity of tryptophan 2,3‐dioxygenase (mainly expressed in the liver), which, together with IDO, is responsible for tryptophan→kynurenine conversion in the periphery ; and (2) changes in brain influx of permeable KP metabolites (eg, tryptophan and kynurenine). Moreover, changes in downstream enzymatic activity (Figure S1) may underestimate or overestimate the calculated kynurenine/tryptophan ratio, and the influx of kynurenine and tryptophan into the brain may have partially masked the activation of the downstream KP steps. Thus, mRNA expression, protein levels, and the activity of brain and peripheral enzymes of the KP after CA/CPR, need to be investigated to better elucidate which branch of the KP is preferentially activated after CA/CPR.

In conclusion, we uphold the involvement of KP in the neurological deficit following CA/CPR, also demonstrating that the degree of peripheral activation of KP may predict functional outcomes of CA. Further studies are needed to clarify the role of each neuroactive kynurenines in the molecular mechanisms underlying the CA‐induced neuropathological events––also considering other downstream metabolites (eg, quinolinic acid)––in order to define KP as a neuroprognostic biomarker after CA/CPR and/or an important pharmacological target to reduce morbidity and mortality after successful CPR.

Sources of Funding

The study was fully supported by the Associazione Amici del Mario Negri.

Disclosures

None.

Data S1

Tables S1–S4

Figures S1–S5

References 96, 97, 98

Click here for additional data file.