High extracellular levels of potassium and trace metals in human brain abscess

Neurochemistry International 82 (2015) 28–32

Contents lists available at ScienceDirect

Neurochemistry International j o u r n a l h o m e p a g e : w w w. e l s e v i e r. c o m / l o c a t e / n c i

High extracellular levels of potassium and trace metals in human brain abscess Daniel Dahlberg a, Jugoslav Ivanovic a, Espen Mariussen b, Bjørnar Hassel b,c,* a b c

Department of Neurosurgery, Oslo University Hospital, Oslo, Norway Norwegian Defence Research Establishment (FFI), Kjeller, Norway Department of Neurology, Oslo University Hospital, Oslo, Norway

A R T I C L E

I N F O

Article history: Received 20 November 2014 Received in revised form 2 February 2015 Accepted 9 February 2015 Available online 12 February 2015 Keywords: Brain abscess Cerebrospinal fluid Epilepsy Serum Potassium Zinc

A B S T R A C T

Brain abscesses frequently cause symptoms such as seizures, delirium, paresis and sensory deficits that could reflect brain edema, increased intracranial pressure, or tissue destruction. However, it is also possible that pus constituents could disturb neuronal function in the surrounding brain tissue. In pus from 16 human brain abscesses, extracellular potassium ([K+]o) was 10.6 ± 4.8 mmol/L (mean ± SD; maximum value 22.0 mmol/L). In cerebrospinal fluid (CSF), [K+]o was 2.7 ± 0.6 mmol/L (N = 14; difference from pus p < 0.001), which is similar to previous control values for [K+]o in CSF and brain parenchyma. Zinc and iron were >40-fold higher in pus than in CSF; calcium, copper, manganese, and chromium were also higher, whereas sodium and magnesium were similar. Pus from 10 extracerebral abscesses (empyemas) also had higher [K+]o, zinc, iron, calcium, copper, manganese, and chromium than did CSF. Brain abscess [K+]o was significantly higher than serum potassium (3.8 ± 0.5 mmol/L; p = 0.0001), indicating that the elevated abscess [K+]o originated from damaged cells (e.g. brain cells and leukocytes), not from serum. High [K+]o could depolarize neurons, high levels of zinc could inhibit glutamate and GABA receptors, and high levels of iron and copper could cause oxidative damage, all of which could contribute to neuronal dysfunction in brain abscess patients. © 2015 Elsevier Ltd. All rights reserved.

1. Introduction Bacterial brain abscesses often cause paresis, sensory disturbance, disorientation, and seizures, symptoms that indicate alterations in neuronal activity (Brouwer et al., 2014; Dahlberg et al., 2014; Kilpatrick, 1997; Sellner and Trinka, 2013; Yang and Zhao, 1993). The causes of such symptoms in brain abscess patients have not been identified. Abscess formation entails focal loss of brain cells, vasogenic edema, increased intracranial pressure, distortion of brain anatomy (Ebeling and Reulen, 1995), opening of the blood–brain barrier (Lo et al., 1994; Yamamoto et al., 1993), and inflammatory responses (Britt et al., 1981), all of which could contribute to the symptoms mentioned above. However, the possibility that pus may contain constituents that affect the function of the surrounding brain tissue has received little attention. Brain abscess formation implies that brain cells die and become replaced by pus. The abscess cavity may therefore represent a mixture of extracellular and intracellular

Abbreviations: [K+]o, extracellular potassium concentration; CSF, cerebrospinal fluid; ROS, reactive oxygen species. * Corresponding author. Norwegian Defence Research Establishment (FFI), 0027 Kjeller, Norway. Tel.: +47 63 80 78 46; fax: +47 63 80 75 09. E-mail address: bjornar.hassel@ffi.no (B. Hassel). http://dx.doi.org/10.1016/j.neuint.2015.02.003 0197-0186/© 2015 Elsevier Ltd. All rights reserved.

fluid compartments, resulting in supraphysiological extracellular concentrations of cations that under normal conditions primarily have an intracellular localization. The continuous migration of polymorphonuclear leukocytes into the abscess cavity (and their subsequent death there) could also contribute to a high extracellular concentration of ‘intracellular’ cations in pus. For example, potassium ions, whose intracellular concentration is normally 135–150 mmol/L (Wright, 2004), could be increased in brain abscesses. Indeed, two previous studies found high extracellular potassium ([K+]o) in skin or dental abscesses with mean [K+]o values of 17 mmol/L (Zimmerli and Gallin, 1988) and 37 mmol/L (Wiese, 1994), respectively; the high [K+]o was thought to have an activating effect on neutrophils in the pus (Zimmerli and Gallin, 1988). A potentially important aspect of [K+]o in brain abscesses is related to the depolarizing effect of high [K+]o on neurons, which could lead to neuronal activation, including seizure activity (Lothman et al., 1975; McBain et al., 1990; Traynelis and Dingledine, 1988) or, at higher [K+]o, to neuronal inactivation through sustained depolarization (Lothman et al., 1975; Obrenovitch and Zilkha, 1995). Therefore, if potassium leaks from the abscess cavity into the surrounding brain tissue, abnormal neuronal activation or inactivation could cause neurological symptoms. Similarly, the extracellular concentration of zinc (Zn) was recently shown to increase after experimental brain infection with Staphylococcus aureus (Hassel et al., 2014). Zn may interfere with normal neurotransmission through

D. Dahlberg et al./Neurochemistry International 82 (2015) 28–32

interaction with GABAA receptors (Buhl et al., 1996; Hosie et al., 2003) and glutamate receptors (Mott et al., 2008; Rachline et al., 2008). The purpose of this study was to determine concentrations of metal ions in brain abscess pus from human patients in order to evaluate a possible contribution from metal ions to the symptoms caused by brain abscesses. We analyzed potassium (K), calcium (Ca), sodium (Na), magnesium (Mg), zinc (Zn), iron (Fe), copper (Cu), manganese (Mn), and chromium (Cr) in the extracellular phase of pus from patients with (intracerebral) brain abscess, reasoning that extracellular cations would be able to diffuse out of the abscess cavity and into the surrounding brain tissue. For comparison we analyzed the cerebrospinal fluid (CSF) from patients undergoing drainage of the cerebral ventricles. To see if changes in extracellular concentrations of cations were specific to intracerebral abscesses we also analyzed pus from extracerebral abscesses (empyemas). 2. Materials and methods 2.1. Patients and samples The Regional Ethics Committee for Medical Research in the southern and eastern part of Norway approved the study, and all participants, or their parents (in the case of children), gave informed, written consent. During the years 2011–2014 pus was collected at surgery from 16 patients with intracerebral, bacterial abscess (median age 60.5 years; range 3–84) and from 10 patients with extracerebral empyema (median age 67 years; range 30–78). Craniotomy was performed in an operating theatre with laminar air flow and filtering of air through high-efficiency particulate arrestance (HEPA)-filters approximately 500 times per hour. Pus, approximately 1–10 mL was evacuated for analysis by aspiration into polypropylene syringes (Sarstedt, Nümbrecht, Germany), which could not be acidrinsed prior to use due to operating theatre procedures. No stainless steel needles or other metal instruments were used to obtain the samples. All pus samples underwent light microscopy. CSF was from 14 patients undergoing external drainage of the cerebral ventricles with an ExactaTM external drainage and monitoring system (Medtronic Neurosurgery, Goleta, CA, USA) as treatment for increased intracranial pressure (median age 52 years; range 1.5– 84). CSF was collected in ExactaTM drainage bags. Intracranial pressure was normal, and the CSF was clear and colorless at the time of sampling. Samples were taken from the bags by allowing CSF to flow into 15 mL polypropylene tubes (Sarstedt) for 2–3 s. Samples were cooled on ice and rapidly centrifuged at 3000 g for 10 minutes at 4 °C. Supernatants were transferred to polypropylene tubes with a 200 μL tip (not acid-washed) equipped with a filter barrier (ART 200, ThermoFisher) and frozen at −70 °C until analysis. Serum, which was used for analysis of K, Na, Mg, Ca (see Section 2.3 below), was obtained after venipuncture with a stainless steel cannula and aspiration of blood into Vacuette® 6 mL serum tubes (Greiner Bio-One, Kremsmünster, Austria). Centrifugation was done after approximately 15 minutes at 1000 g and 4 °C for 10 minutes to obtain serum. Prior to surgery, all brain abscess patients underwent CT or MRI scan of the brain. Volumes of abscesses were calculated from the radii in the x, y, and z planes, using the ellipsoid formula (4/3 π · rx · ry · rz; Dahlberg et al., 2014; Mistry et al., 2013).

29

a 2 mL polypropylene syringe (Sarstedt) and transferred to 15 mL polypropylene tubes (N = 15) by a neurosurgeon dressed for surgery. Also present were another neurosurgeon, an anesthesiologist, a neurologist, and three nurses. Another set of blanks (N = 10) was prepared by filling ExactaTM CSF drainage bags with Milli-Q water (400 mL per bag). From the bags water was allowed to flow into 15 mL polypropylene tubes (Sarstedt) for 2–3 s. A series of blanks (N = 5) was run to assess any contaminating effect of pipetting. Ultrapure HNO3, 4.950 mL, 0.65%, were pipetted into 15 mL polypropylene tubes with 5 mL polypropylene tips (Finntip, ThermoFisher, Waltham, MA USA) that had been flushed with ultrapure HNO3, 0.65% (as was the case when patient samples were prepared). Then 50 μL ultrapure HNO3, 0.65%, were added with a pipette, each time using a new, unflushed 200 μL tip equipped with a filter barrier (ART 200, ThermoFisher). These blanks were compared to ultrapure HNO3, 5 mL, 0.65%, which was decanted (without pipetting) into 15 mL polypropylene tubes. To control for background contamination during analysis another set of blanks, ultrapure HNO3, 0.65%, (N = 16) was analyzed between samples. The mean values of these blanks were subtracted from sample and field blank values. Detection limits were calculated as 3 × SD values obtained with these blanks. 2.3. Measurement of K, Na, Mg, and Ca For measurements of metal ions in supernatants of pus and CSF, samples were diluted 1:100 in 5 mL ultrapure HNO3, 0.65% (vol/ vol; Merck; distilled in-house). K, Na, Mg, Ca, Zn, Fe, Cu, Mn, and Cr were analyzed on an inductively coupled plasma mass spectrometer (ICP–MS; Thermo X-series II ICP-MS; ThermoFisher Scientific). An internal concentration standard containing 45Sc (scandium), 103Rh (rhodium), 115In (indium), and 175Lu (lutetium) was added to each sample to correct for any drift in signal intensity. The elements of interest were quantified with the use of a four point standard curve (1–1000 ppm for K, Na, Mg, Ca and 0.1–100 ppm for Zn, Fe, Cu, Mn, and Cr). Reference solutions (Rain-97 or Bigmoose02, and Battle-02, TM 23.4, TMDA 61.2, and TMDA 53.3; Analytical Reference Material, Environment Canada, Gatineau, Quebec, Canada) were analyzed in addition to in-house made standards. Serum was analyzed with respect to K, Na, Mg, and Ca by indirect potentiometry and ion-selective electrodes (Modular P800, Roche; Basel, Switzerland). Values were obtained on the same day as, but prior to, surgery (for abscess or empyema) or sampling of CSF. 2.4. Data presentation and statistics Data on cations are given as means ± SD values. Values for K, Na, Mg, Ca, and trace metals were not normally distributed by the Kolmogorov–Smirnov test or the D’Agostino and Pearson’s test and were analyzed non-parametrically by Kruskal–Wallis’ and Dunn’s tests. Age and time from symptom onset to surgery are given as median values with full range. Correlations were analyzed with Spearman’s tests. All data sets were tested with 1-sample tests for difference from zero. Statistical testing was done using the Prism 5 statistics program (GraphPad, Software, La Jolla, CA, USA). A p-value <0.05 was considered statistically significant.

2.2. Blanks 3. Results Blanks were water from a Milli-Q® Advantage A10 ultrapure water purification system (Milli-Q water; Merck Millipore, Darmstadt, Germany) equipped with a Quantum® Tex1 polishing cartridge for removal of metal ions. During surgery in the operating theatre, a 50 mL polypropylene tube (Sarstedt) containing Milli-Q water was opened close to the head of a patient, and water was drawn with

3.1. Blank values Blanks taken in the operating theatre with patient, surgeons, anesthesiologist, and nurses present contained significantly higher levels of K, Na, Ca, and Mg than blanks that had not been handled

30

D. Dahlberg et al./Neurochemistry International 82 (2015) 28–32

Table 1 Extracellular concentrations of Zn, Fe, Cu, Mn, and Cr in CSF, brain abscess pus, and empyemas together with corresponding blank values. CSF was from 14 patients with external cerebroventricular drains; pus was from 16 patients with brain abscess and from 10 patients with empyema. ‘CSF blanks’ were from CSF drainage bags filled with ultrapure Milli-Q water (N = 10). ‘Abscess blanks’ was Milli-Q water sampled in the operating theatre during brain surgery (N = 15). From all values are subtracted values obtained with Milli-Q water blanks (N = 16) run between samples. The detection limit for the ICP–MS was calculated as 3× the SD obtained with the latter blanks. Data are μmol/L; mean ± SD values. Cation

CSF

CSF blanks

Brain abscess

Empyema

Abscess blanks

3 × SD

Zn, μmol/L Fe, μmol/L Cu, μmol/L Mn, μmol/L Cr, μmol/L

1.1 ± 0.8 2.0 ± 1.2 1.5 ± 1.4 0.032 ± 0.027 0.049 ± 0.031

0.6 ± 0.1 0.00 ± 0.07$ 0.03 ± 0.04$ 0.003 ± 0.005$ 0.003 ± 0.001$

42 ± 18*** 270 ± 220*** 7.5 ± 5.3*** 0.39 ± 0.27*** 0.60 ± 0.47***

32 ± 14*** 470 ± 450*** 14 ± 5.0*** 0.33 ± 0.20*** 0.53 ± 0.37***

0.22 ± 0.17 0.2 ± 0.5$ 0.07 ± 0.1 0.014 ± 0.013 0.011 ± 0.013$

0.11 0.35 0.016 0.007 0.017

Asterisk: Different from CSF values, ***p < 0.001; Kruskal–Wallis test with Dunn’s correction for multiple comparisons. $ : not significantly different from zero.

1982; Romarís et al., 2011) and for the extracellular fluid of the brain in experimental animals (Heinemann et al., 1977; Windmüller et al., 2005). The extracellular concentration of Ca was significantly higher in both brain abscesses and empyemas than in CSF (Fig. 1b). However, free (ionized) calcium was not determined in pus, so it is not known to what degree Ca in pus was protein bound. Extracellular Na and Mg did not differ significantly between pus and CSF (Fig. 1c, d).

Extracellular potassium

a

b

***

10

***

8

Extracellular calcium

4

6 4

3

mmol Ca2+/L

12

mmol K+/L

in the operating theatre (p < 0.05; t-test), but values never exceeded 10 μmol/L. The same was true for blanks taken from CSF drainage bags; these blanks had higher levels of K, Ca, Na, and Mg than Milli-Q water (p < 0.05; t-test), but they did not exceed 10 μmol/ L. The use of pipettes with unwashed tips added an extra 1.1 μmol Ca/L and 0.4 μmol Mg/L; other metal ions were not introduced with the use of pipettes. None of the blanks (from the operating theatre or CSF drainage bags) had levels of trace metals above the levels found in Milli-Q water (see Table 1: ‘CSF blanks’ and ‘Abscess blanks’ for trace metal levels in blanks). However, CSF blanks (from CSF drainage bags) had levels of Zn that were not significantly different from those found in CSF samples (Table 1). Some blank values were not significantly different from zero (1-sample test; Table 1). Detection limits for K, Ca, Na, and Mg were 0.007, 0.002, 0.009, and 0.0005 mmol/L, respectively. Detection limits for trace metals are given in Table 1. The mean values for blanks run between samples that were subtracted from samples and field blank values were 0 μmol K/L, 2 μmol Ca/L, 5 μmol Na/L, 0.2 μmol Mg/L, 0 nmol Zn/L, 0.28 μmol Fe/L, 9 nmol Cu/L, 3 nmol Mn/L, and 0.3 nmol Cr/L.

***

***

2 1

2

3.2. [K+]o in CSF and pus

0

0 CSF

Abscess

CSF

Empyema

Abscess

Empyema

+

3.3. Ca, Na, and Mg in CSF and pus Mean CSF values for Ca, Na, and Mg were 1.2 mmol/L, 145 mmol/ L, and 1.1 mmol/L, respectively, which are in excellent agreement with values reported previously for CSF control samples (Bogden et al., 1977; Bradbury et al., 1963; Mercieri et al., 2012; Pye and Aber,

c

d

Extracellular sodium

200

2.0

150

1.5

mmol Mg2+/L

mmol Na+/L

100 50 0 CSF

Abscess

5 4 3 2 1 0

Empyema

Extracellular magnesium

1.0 0.5 0.0 CSF

Abscess

Empyema

Serum values

e mmol/L

In CSF from the cerebral ventricles, mean [K ]o was 2.7 mmol/L (Fig. 1a), which is in excellent agreement with values obtained in control samples of CSF in previous studies (Bradbury et al., 1963; Mercieri et al., 2012; Pye and Aber, 1982) and for the extracellular fluid of the brain of experimental animals (Heinemann et al., 1977; Somjen, 1979). In pus from intracerebral abscesses, [K+]o was significantly higher, with a mean value of 10.6 mmol/L and a maximum value of 22.0 mmol/L (Fig. 1a). In empyemas, mean [K + ] o was 7.1 mmol/L, which also was significantly higher than in CSF (Fig. 1a). Median time from symptom onset to surgery for brain abscess was 9 days (range 4–56). There was no correlation between [K+]o in pus and number of days from symptom onset to surgery (r = −0.07; p = 0.8). The median volume of the brain abscesses was 14 cm3 (range 2–45). There was no correlation between [K+]o and abscess volume and (r = 0.11; p = 0.69). Median time from symptom onset to surgery for empyema was 14 days (range 5–84). There was no significant correlation between [K+]o in pus and number of days from symptom onset to surgery (r = −0.32; p = 0.37). All samples from brain abscesses and empyemas were rich in polymorphonuclear leukocytes, confirming the presence of pus.

K+

total Ca2+

ion. Ca2+

Na+:50

Mg2+

Fig. 1. Extracellular concentration of K, Na, Mg, and Ca in CSF, brain abscess, empyema, and serum. CSF was from 14 patients with external cerebroventricular drains, pus was from 16 patients with brain abscess and from 10 patients with empyema. Samples were rapidly cooled and centrifuged to obtain the extracellular phase. a: extracellular potassium ([K+]o), b: extracellular calcium, c: extracellular sodium, d: extracellular magnesium, e: serum values (serum sodium values are divided by 50). White columns: CSF; black columns: brain abscess; grey columns: empyema. Data are mmol/L; mean + SD values. Asterisks: different from CSF values, ***: p < 0.001; Kruskal– Wallis test with Dunn’s correction for multiple comparisons.

D. Dahlberg et al./Neurochemistry International 82 (2015) 28–32

3.4. Serum K, Ca, Na, and Mg We next investigated whether pus values of K, Ca, Na, or Mg might reflect the serum levels of these cations. The patient groups did not differ with respect to serum levels of K, Ca (total or ionized), Na, or Mg (Fig. 1e). [K+]o in brain abscesses was 157 ± 117% higher than serum K (p = 0.0001). Similarly, [K+]o in empyemas was 85 ± 95% higher than serum K (p = 0.027). In contrast, [K+]o in CSF was 29 ± 23% lower than serum K (p = 0.0020). In patients with brain abscess or empyema, total Ca in pus was similar to total serum Ca (compare Fig. 1b and e). However, ionized serum Ca was 50 ± 5% of total serum Ca in these patients and was lower than total Ca in pus (p < 0.01). In patients undergoing drainage of the cerebral ventricles Ca in CSF was 53 ± 16% of total serum Ca (p = 6 × 10−6), but ionized serum Ca, being 55 ± 3% of total serum Ca, was not different from Ca in CSF. Serum Na was not significantly different from corresponding Na values in pus or CSF. In patients with brain abscess, Mg in pus was 30 ± 25% higher than serum Mg (p = 0.012). This was not the case in empyemas, where Mg in pus was 121 ± 40% of serum Mg. CSF Mg was 63 ± 45% higher than serum Mg (p = 0.006). 3.5. Trace metal ions in CSF and pus Mean CSF values for Fe (2.0 μmol/L), Cu (1.5 μmol/L), Mn (0.032 μmol/L), and Cr (0.049 μmol/L) were in broad agreement with previously published values for control CSF samples (Bogden et al., 1977; Boström et al., 2009; Gellein et al., 2008; Michalke and Nischwitz, 2010; Roos et al., 2013). This was true even for Zn (1.1 μmol/L), but the concentration of Zn in CSF samples was not significantly different from concentrations found in blank samples from CSF drainage bags (Table 1). All trace metal data sets from clinical samples were greater than zero (1-sample test). All trace metals were present at higher concentrations in brain abscesses and empyemas than in corresponding blank samples (Table 1). Zn, Fe, and Cu were present at >100 times the level in blank samples, whereas Mn and Cr were present at >20 times the level in blank samples, as could be calculated from the mean levels. Further, all trace metals were present at significantly higher concentrations in brain abscesses and empyemas than in CSF (Table 1). The mean concentrations of Zn and Fe were >40 times higher in brain abscesses than in CSF, and the mean concentrations of Cu, Mn, and Cr were also higher in brain abscesses than in CSF. The concentrations of trace metals in empyemas were similar those in brain abscesses (Table 1). 4. Discussion 4.1. Elevated [K+]o in brain abscesses and empyemas We show here that the extracellular phase of pus from brain abscesses has elevated [K+]o. The pus concentration of K was almost four times higher than [K+]o in CSF, as could be calculated from the mean values, and the highest value measured was 22 mmol/L. The source of the elevated [K+]o in brain abscesses was in all likelihood the intracellular pool of K from cells (brain cells or leukocytes) undergoing necrosis, not serum K. This follows from the observation that [K+]o in abscesses was higher than in serum, so that even though a brain abscess entails leakage of serum constituents through the blood–brain barrier (Lo et al., 1994; Yamamoto et al., 1993), the circulation could not have been the source of the elevated [K+]o in brain abscesses. The high [K+]o in pus was not specific to brain abscesses, but was also found in empyemas, which are situated outside the brain proper. Brain cells would therefore not be a likely source of high [K+]o in empyemas, but dying leukocytes could be such a source. A high [K+]o

31

has been reported previously in skin and dental abscesses (Wiese, 1994; Zimmerli and Gallin, 1988), underscoring high [K+]o as a likely general feature of bacterial abscesses. 4.2. A role for [K+]o in neurological dysfunction in brain abscess patients? An important aspect of a high [K+]o is its depolarizing effect on neurons. A brain abscess is surrounded by a capsule composed of fibrous tissue and activated astrocytes (Britt et al., 1981), but this capsule does not constitute a diffusion barrier (Lo et al., 1994; Yamamoto et al., 1993), so diffusion of K from the abscess cavity into the surrounding brain tissue may occur. An elevated [K+]o may cause both neuronal activation and inactivation. Experimentally, a [K+]o of 8–10 mmol/L produces seizure activity in both the cat brain and brain slices (Lothman et al., 1975; McBain et al., 1990; Traynelis and Dingledine, 1988), and a [K+]o of approximately 10 mmol/L is regularly seen in the brains of experimental animals with ongoing seizure activity (Moody et al., 1974; Somjen, 1979). In contrast, a [K+]o in excess of 12–15 mmol/L tends to cause neuronal inactivation through continuous depolarization (Lothman et al., 1975; Obrenovitch and Zilkha, 1995). Clearly, both effects of high [K+]o could, in conjunction with other consequences of brain abscess (see Section 1), contribute to cerebral symptoms, including seizures, delirium, and, depending on abscess localization, paresis and sensory deficits, which are common symptoms in brain abscess patients (Brouwer et al., 2014; Dahlberg et al., 2014; Kilpatrick, 1997; Sellner and Trinka, 2013; Yang and Zhao, 1993). The depolarizing effect of high [K+]o could potentiate the excitatory effect of high concentrations of glutamate and aspartate, which was recently reported in brain abscess pus (Dahlberg et al., 2014). 4.3. Elevated levels of Ca and trace metals in brain abscesses and empyemas Compared to the levels in CSF, the level of Ca and all measured trace metals was higher in brain abscesses and empyemas. However, because CSF values for some trace elements (especially Zn) were similar to values measured in blanks, and because CSF was from patients who were treated for increased intracranial pressure, these data should not be taken to reflect normal CSF values. We do not know to which degree cations in pus were free or protein-bound, but in a recent study we found an increase in extracellular free Zn after experimental brain infection with Staphylococcus aureus (Hassel et al., 2014). Therefore, it is possible that for instance Zn may diffuse from the abscess cavity and into the surrounding brain tissue, where it could affect both inhibitory and excitatory neurotransmission through its inhibitory activity at GABAA receptors (Buhl et al., 1996; Hosie et al., 2003) and glutamate receptors (Mott et al., 2008; Rachline et al., 2008). The high levels of Fe and Cu in pus may facilitate formation of reactive oxygen species (ROS; Halliwell and Gutteridge, 1984; Kozlowski et al., 2014). ROS formation is thought to be the cause of the seizure-precipitating effect of elevated levels of free iron in the brain (Willmore et al., 1978, 1986). ROS formation is also thought to be the cause of the pro-apoptotic effect of Cu on neurons (Chen et al., 2009) and its pro-necrotic effect in astrocytes (Scheiber et al., 2010) at concentrations seen in pus in the present study. Formation of ROS may also be stimulated by Mn (Stephenson et al., 2013) and Cr (Bagchi et al., 2002; Davidson et al., 2007) at low micromolar concentration, but the concentration of these ions in pus in the present study was submicromolar, so it is difficult to evaluate their importance for symptom generation in our patients. Increased levels of Zn, Fe, Cu, and Mn have been reported in pus from other locations than the brain previously (Bryant et al., 2004; Domej et al., 2000), pointing to elevated levels of trace metals as a

32

D. Dahlberg et al./Neurochemistry International 82 (2015) 28–32

common in pus. The special relevance of high levels of trace metals in brain abscesses lies in their effects on neuronal and glial activity and integrity (Buhl et al., 1996; Chen et al., 2009; Hosie et al., 2003; Mott et al., 2008; Rachline et al., 2008; Scheiber et al., 2010; Willmore et al., 1978, 1986) and this points to the importance of pus evacuation. These ions may act in conjunction with other pus constituents, such as high levels of glutamate and aspartate (Dahlberg et al., 2014), as well as with other consequences of an abscess, including brain edema, increased intracranial pressure, and distortion of brain anatomy, to cause neurological symptoms in brain abscess patients. Acknowledgements This study was supported by The Norwegian Epilepsy Society and The Norwegian Health Association (Grant #1513). The sponsors did not have any role in the study design, data collection, analysis, or interpretation, or in the writing of this manuscript or the decision to submit it for publication. D.D. and B.H. designed the study. All authors took part in sample preparation, data interpretation, and manuscript preparation, and all authors have read and approved this manuscript. References Bagchi, D., Stohs, S.J., Downs, B.W., Bagchi, M., Preuss, H.G., 2002. Cytotoxicity and oxidative mechanisms of different forms of chromium. Toxicology 180, 5–22. Bogden, J.D., Troiano, R.A., Joselow, M.M., 1977. Copper, zinc, magnesium, and calcium in plasma and cerebrospinal fluid of patients with neurological diseases. Clin. Chem. 23, 485–489. Boström, F., Hansson, O., Gerhardsson, L., Lundh, T., Minthon, L., Stomrud, E., et al., 2009. CSF Mg and Ca as diagnostic markers for dementia with Lewy bodies. Neurobiol. Aging 30, 1265–1271. Bradbury, M.W., Stubbs, J., Hughes, I.E., Parker, P., 1963. The distribution of potassium, sodium, chloride and urea between lumbar cerebrospinal fluid and blood serum in human subjects. Clin. Sci. 25, 97–105. Britt, R.H., Enzmann, D.R., Yeager, A.S., 1981. Neuropathological and computerized tomographic findings in experimental brain abscess. J. Neurosurg. 55, 590–603. Brouwer, M.C., Coutinho, J.M., van de Beek, D., 2014. Clinical characteristics and outcome of brain abscess: systematic review and meta-analysis. Neurology 82, 806–813. Bryant, R.E., Crouse, R., Deagen, J.T., 2004. Zinc, iron, copper, selenium, lactoferrin, and ferritin in human pus. Am. J. Med. Sci. 327, 73–76. Buhl, E.H., Otis, T.S., Mody, I., 1996. Zinc-induced collapse of augmented inhibition by GABA in a temporal lobe epilepsy model. Science 271, 369–373. Chen, X., Lan, X., Mo, S., Qin, J., Li, W., Liu, P., et al., 2009. p38 and ERK, but not JNK, are involved in copper-induced apoptosis in cultured cerebellar granule neurons. Biochem. Biophys. Res. Commun. 379, 944–948. Dahlberg, D., Ivanovic, J., Hassel, B., 2014. High extracellular concentration of excitatory amino acids glutamate and aspartate in human brain abscess. Neurochem. Int. 69, 41–47. Davidson, T., Ke, Q., Costa, M., 2007. Selected molecular mechanisms of metal toxicity and carcinogenicity. In: Nordberg, G.F., Fowler, B.A., Nordberg, M., Friberg, L.T. (Eds.), Handbook on the Toxicology of Metals, third ed. Elsevier., Amsterdam, pp. 79–100. Domej, W., Krachler, M., Goessler, W., Maier, A., Irgolic, K.J., Lang, J.K., 2000. Concentrations of copper, zinc, manganese, rubidium, and magnesium in thoracic empyemata and corresponding sera. Biol. Trace Elem. Res. 78, 53–66. Ebeling, U., Reulen, H.J., 1995. Space-occupying lesions of the sensori-motor region. Adv. Tech. Stand. Neurosurg. 22, 137–181. Gellein, K., Skogholt, J.H., Aaseth, J., Thoresen, G.B., Lierhagen, S., Steinnes, E., et al., 2008. Trace elements in cerebrospinal fluid and blood from patients with a rare progressive central and peripheral demyelinating disease. J. Neurol. Sci. 266, 70–78. Halliwell, B., Gutteridge, J.M., 1984. Oxygen toxicity, oxygen radicals, transition metals and disease. Biochem. J. 219, 1–14. Hassel, B., Dahlberg, D., Mariussen, E., Goverud, I.L., Antal, E.A., Tønjum, T., et al., 2014. Brain infection with Staphylococcus aureus leads to high extracellular levels of glutamate, aspartate, γ-aminobutyric acid, and zinc. J. Neurosci. Res. 92, 1792– 1800.

Heinemann, U., Lux, H.D., Gutnick, M.J., 1977. Extracellular free calcium and potassium during paroxsmal activity in the cerebral cortex of the cat. Exp. Brain Res. 27, 237–243. Hosie, A.M., Dunne, E.L., Harvey, R.J., Smart, T.G., 2003. Zinc-mediated inhibition of GABAA receptors: discrete binding sites underlie subtype specificity. Nat. Neurosci. 6, 362–369. Kilpatrick, C., 1997. Epilepsy and brain abscess. J. Clin. Neurosci. 4, 26–28. Kozlowski, H., Kolkowska, P., Watly, J., Krzywoszynska, K., Potocki, S., 2014. General aspects of metal toxicity. Curr. Med. Chem. 21, 3721–3740. Lo, W.D., Wolny, A., Boesel, C., 1994. Blood-brain barrier permeability in staphylococcal cerebritis and early brain abscess. J. Neurosurg. 80, 897–905. Lothman, E., Lamanna, J., Cordingley, G., Rosenthal, M., Somjen, G., 1975. Responses of electrical potential, potassium levels, and oxidative metabolic activity of the cerebral neocortex of cats. Brain Res. 88, 15–36. McBain, C.J., Traynelis, S.F., Dingledine, R., 1990. Regional variation of extracellular space in the hippocampus. Science 249, 674–677. Mercieri, M., De Blasi, R.A., Palmisani, S., Forte, S., Cardelli, P., Romano, R., et al., 2012. Changes in cerebrospinal fluid magnesium levels in patients undergoing spinal anaesthesia for hip arthroplasty: does intravenous infusion of magnesium sulphate make any difference? A prospective, randomized, controlled study. Br. J. Anaesth. 109, 208–215. Michalke, B., Nischwitz, V., 2010. Review on metal speciation analysis in cerebrospinal fluid-current methods and results: a review. Anal. Chim. Acta 682, 23–36. Mistry, R.D., Marin, J.R., Alpern, E.R., 2013. Abscess volume and ultrasound characteristics of community-associated methicillin-resistant Staphylococcus aureus infection. Pediatr. Emerg. Care 29, 140–144. Moody, W.J., Futamachi, K.J., Prince, D.A., 1974. Extracellular potassium activity during epileptogenesis. Exp. Neurol. 42, 248–263. Mott, D.D., Benveniste, M., Dingledine, R.J., 2008. pH-dependent inhibition of kainate receptors by zinc. J. Neurosci. 28, 1659–1671. Obrenovitch, T.P., Zilkha, E., 1995. High extracellular potassium, and not extracellular glutamate, is required for the propagation of spreading depression. J. Neurophysiol. 73, 2107–2114. Pye, I.F., Aber, G.M., 1982. Interrelations between cerebrospinal fluid and plasma inorganic ions and glucose in patients with chronic renal failure. J. Clin. Pathol. 35, 631–637. Rachline, J., Perin-Dureau, F., Le Goff, A., Neyton, J., Paoletti, P., 2008. The micromolar zinc-binding domain on the NMDA receptor subunit NR2B. J. Neurosci. 25, 308–317. Romarís, E.M., Cervantes, I.I., López, J.M., Marcén, J.F., 2011. Concentration of calcium and magnesium and trace elements (zinc, copper, iron and manganese) in cerebrospinal fluid: a try of a pathophysiological classification. J. Trace Elem. Med. Biol. 25 (Suppl. 1), S45–S49. Roos, P.M., Vesterberg, O., Syversen, T., Flaten, T.P., Nordberg, M., 2013. Metal concentrations in cerebrospinal fluid and blood plasma from patients with amyotrophic lateral sclerosis. Biol. Trace Elem. Res. 151, 159–170. Scheiber, I.F., Schmidt, M.M., Dringen, R., 2010. Zinc prevents the copper-induced damage of cultured astrocytes. Neurochem. Int. 57, 314–322. Sellner, J., Trinka, E., 2013. Clinical characteristics, risk factors and pre-surgical evaluation of post-infectious epilepsy. Eur. J. Neurol. 20, 429–439. Somjen, G.G., 1979. Extracellular potassium in the mammalian central nervous system. Annu. Rev. Physiol. 41, 159–177. Stephenson, A.P., Schneider, J.A., Nelson, B.C., Atha, D.H., Jain, A., Soliman, K.F., et al., 2013. Manganese-induced oxidative DNA damage in neuronal SH-SY5Y cells: attenuation of thymine base lesions by glutathione and N-acetylcysteine. Toxicol. Lett. 218, 299–307. Traynelis, S.F., Dingledine, R., 1988. Potassium-induced spontaneous electrographic seizures in the rat hippocampal slice. J. Neurophysiol. 59, 259–276. Wiese, K.G., 1994. Electrolyte concentration, real and osmotic pressure in abscesses. Zentralbl. Chir. 119, 54–59. Willmore, L.J., Sypert, G.W., Munson, J.V., Hurd, R.W., 1978. Chronic focal epileptiform discharges induced by injection of iron into rat and cat cortex. Science 200, 1501–1503. Willmore, L.J., Triggs, W.J., Gray, J.D., 1986. The role of iron-induced hippocampal peroxidation in acute epileptogenesis. Brain Res. 382, 422–426. Windmüller, O., Lindauer, U., Foddis, M., Einhäupl, K.M., Dirnagl, U., Heinemann, U., et al., 2005. Ion changes in spreading ischaemia induce rat middle cerebral artery constriction in the absence of NO. Brain 128, 2042–2051. Wright, S.H., 2004. Generation of resting membrane potential. Adv. Physiol. Educ. 28, 139–142. Yamamoto, M., Jimbo, M., Ide, M., Tanaka, N., Umebara, Y., Hagiwara, S., 1993. Penetration of intravenous antibiotics into brain abscesses. Neurosurgery 33, 44–49. Yang, S.Y., Zhao, C.S., 1993. Review of 140 patients with brain abscess. Surg. Neurol. 39, 290–296. Zimmerli, W., Gallin, J.I., 1988. Pus potassium. Inflammation 12, 37–43.