SHORT COMMUNICATION

Salivary a-amylase response to endotoxinadministration in humans

Jan-Sebastian Grigoleit a,*, Jennifer S. Kullmann a,b, Reiner Oberbeck c,Manfred Schedlowski a, Harald Engler a

a Institute of Medical Psychology and Behavioral Immunobiology, University Hospital Essen, University of Duisburg-Essen, 45122Essen, Germanyb Institute of Diagnostic and Interventional Radiology and Neuroradiology, University Hospital Essen, 45122 Essen, GermanycDepartment of Trauma Surgery, University Hospital Essen, 45122 Essen, Germany

Received 3 September 2012; received in revised form 4 December 2012; accepted 9 January 2013

Psychoneuroendocrinology (2013) 38, 1819—1823

KEYWORDSAmylase;Inflammation;Endotoxin;Lipopolysaccharide;Norepinephrine;AutonomicNervous System

Summary Salivary a-amylase (sAA) is a digestive enzyme that plays also an important role inmucosal immunity. Secretion of the sAA is largely under the control of the autonomic nervoussystem and increases in sAA activity have repeatedly been observed in response to variousstressors. The present study aimed at investigating whether and to what extent sAA activity levelsare affected during systemic inflammation. Fourteen healthy male volunteers received intrave-nous injections of either bacterial endotoxin or placebo at two different occasions in a random-ized and double-blinded manner. sAA activity was monitored over a period of 6 h together withinflammatory markers, plasma norepinephrine (NE) and salivary cortisol levels, vital parameters,and state anxiety. Endotoxin administration elicited a transient inflammatory response reflectedby increases in body temperature, whole blood cell counts, and circulating levels of interleukin(IL)-6. The immune changes were accompanied by a transient increase in sAA activity, elevationsin salivary cortisol and plasma NE concentrations, as well as increases in heart rate and stateanxiety. Although sAA and plasma NE responses showed distinct time courses, a significantpositive correlation over the total observation period was found. Whether the observed sAAresponse is driven by an increase in sympathetic activity or more generally reflects inflammationinduced changes in sympathetic-parasympathetic balance remains to be elucidated.# 2013 Elsevier Ltd. All rights reserved.

Available online at www.sciencedirect.com

j our na l h omepa g e: www.e l se v ie r.c om/l oca te/ psyne ue n

* Corresponding author at: Institute of Medical Psychology andBehavioral Immunobiology, University Hospital Essen, University ofDuisburg-Essen, Hufelandstrasse 55, D-45122 Essen, Germany.Tel.: +49 201 723 83497; fax: +49 201 723 5948.

E-mail address: jan-sebastian.grigoleit@uk-essen.de(J.-S. Grigoleit).

0306-4530/$ — see front matter # 2013 Elsevier Ltd. All rights reservehttp://dx.doi.org/10.1016/j.psyneuen.2013.01.003

1. Introduction

Salivary a-amylase (sAA) is a digestive enzyme responsible forthe breakdown of starch in the oral cavity. Besides its diges-tive function, it also plays an important role in mucosalimmunity, as it inhibits bacterial growth and adherence(Scannapieco et al., 1993). Together with other proteins,

d.

1820 J.-S. Grigoleit et al.

sAA is synthesized in the acinar cells of the salivary glands,and its secretion into oral fluids is largely controlled by theautonomic nervous system (ANS): sympathetic nerves mainlyregulate salivary protein secretion, whereas parasympa-thetic nerves mainly control salivary flow rate (Proctor andCarpenter, 2007).

Numerous studies in humans and animals have demon-strated that sAA protein or activity levels rapidly increasefollowing administration of sympathomimetic drugs as well asin response to physiological and psychological stressors(Nater and Rohleder, 2009). Furthermore, administrationof the b-adrenoceptor antagonist propranolol was shownto abrogate stress induced increases in sAA activity (vanStegeren et al., 2006). Together with the observation thatsAA activity concomitantly increased with plasma norepi-nephrine (NE) and epinephrine levels under various stressconditions, this led to the assumption that sAA might be apromising non-invasive marker to measure sympathetic acti-vation. However, this view has recently been questioned,since stress induced alterations in sAA activity do not or onlymoderately correlate with other SNS markers such as cardiacpre-ejection period, skin conductance, and plasma NE (Boschet al., 2011).

Systemic inflammation, as a result of bacterial infection ortissue injury, is a potent stimulus for the ANS. Pro-inflamma-tory cytokines such as interleukin (IL)-6 and tumor necrosisfactor (TNF)-a that are released by immune cells into circu-lation activate through neural and humoral afferent path-ways autonomic centers in the brain. This leads to apronounced rise in central and peripheral NE levels (Engleret al., 2011; Grigoleit et al., 2011) as well as increased vagalefferent activity (Tracey, 2002). The combined activation ofthe sympathetic and parasympathetic systems modulates viaadrenergic and cholinergic pathways the activity of inflam-matory cells and thereby provides an important regulatoryfeedback on the immune system (Meisel et al., 2005).

Although it is well-known that the secretion of salivaryproteins is largely under autonomic control, the effect ofsystemic inflammation on sAA levels has not been studied yet.Thus, using a human endotoxemia model, we investigatedhere whether and to what extent sAA activity levels areaffected during the course of an acute experimental inflam-mation. Healthy male volunteers received in a randomizedand double-blinded manner intravenous injections of either alow dose of bacterial endotoxin (lipopolysaccharide, LPS) orplacebo (saline) at two different occasions. sAA activity wasmonitored over a period of 6 h together with inflammatorymarkers, plasma NE and salivary cortisol levels, vital para-meters, and state anxiety. Correlation analyses were per-formed to examine potential associations between sAAactivity and plasma NE levels.

2. Methods

2.1. Participants

Fourteen healthy male volunteers were enrolled in this study.The sample represents a subset of subjects from a largerstudy (n = 34) investigating dose-dependent effects of endo-toxin on neurobehavioral functions (Grigoleit et al., 2011).Participants underwent extensive physical and psychiatric

screenings, and were not allowed to smoke, to drink alcohol,or to take any kind of medication during and at least sevendays prior to the study. One subject had to be excluded fromthe analyses due to an extraordinarily high baseline NE level(Grubbs’ test for outliers, p < 0.01). Thus, the final studyconsisted of thirteen subjects with a mean age of 26.2 � 3.6years (range: 20—33) and a mean body mass index of24.0 � 2.4 (range: 20.3—29.6). Participants were informedabout the study design and written informed consent wasobtained. All procedures were approved by the Ethics Com-mittee of the University of Duisburg-Essen and were carriedout in accordance with the Declaration of Helsinki.

2.2. Experimental design

Each subject received, in a randomized and double-blindedmanner, an intravenous injection of either 0.4 ng LPS/kg ofbody weight (Escherichia coli, Lot G3E069; US Pharmacopeia,Rockville, MD) or placebo (saline) at two different occasionswith an interval of 14 days between treatments. The experi-ments were conducted in a medically equipped room andwere supervised by emergency physicians. Details of theexperimental procedure and preparation of the LPS havebeen previously described (Grigoleit et al., 2011). Briefly,an intravenous catheter was inserted into an antecubitalforearm vein for repeated blood collection and LPS/placeboapplication. After a rest period of 20 min, body temperatureas well as heart rate and blood pressure were measured withan auricular thermometer and a sphygmomanometer, respec-tively, and a first blood and saliva sample was obtained.Thirty minutes later, the injection was performed. Assess-ment of vital signs as well as saliva and blood collection tookplace at 1, 2, 3, 4, and 6 h after the injection. After the lastblood draw, subjects were discharged and asked to return fora final medical checkup on the following day.

2.3. Biochemical analyses and questionnaires

White blood cell (WBC) counts were determined in EDTA-treated blood samples using an automated cell counter(Sysmex KX-21N, Norderstedt, Germany). Plasma for endo-crine and cytokine analyses was obtained by centrifugationand was stored at �80 8C until analysis. Saliva was obtainedby chewing the synthetic swab of a commercially availablecollection device (Salivette; Sarstedt, Numbrecht, Germany)for exactly 1 min. Subjects were not allowed to eat and drinkat least for 20 min prior to saliva collection. Saliva wasdrained from the swab by centrifugation and stored at�20 8C. Plasma IL-6 levels were quantified using a bead-based assay (Bio-Plex Pro Human Cytokine Assay; Bio-RadLaboratories, Munich, Germany). Samples were preparedaccording to the manufacturer’s instructions and were ana-lyzed on a FACSCanto II flow cytometer using FACSDiva soft-ware (BD Immunocytometry Systems, Heidelberg, Germany).Absolute cytokine levels were calculated based on the meanfluorescence intensity of IL-6 standard dilutions with a 4-parameter logistic curve model. The detection limit of the IL-6 assay was 0.2 pg/ml. Plasma levels of NE and salivary levelsof cortisol were determined using enzyme-linked immuno-sorbent assays (Noradrenaline ELISA; Labor DiagnostikaNord, Nordhorn, Germany; Cortisol ELISA; IBL International,

Salivary a-amylase response to endotoxin administration in humans 1821

Hamburg, Germany) according to the manufacturer’s instruc-tions. The detection limits of the assays were 0.26 nmol/l(NE) and 0.83 nmol/l (cortisol), respectively. Salivary a-amy-lase activity was determined using an enzymatic assay (Sali-vary Alpha-Amylase Assay Kit; Salimetrics Europe, Suffolk,

Figure 1 Body temperature (A), white blood cell (WBC) counts (Bnorepinephrine levels (E), salivary cortisol levels (F), heart rate (G),

and after intravenous injection (arrow) of LPS (0.4 ng/kg body weightstudy days. Data are presented as means � SEM. Asterisks indicate re***p < 0.001).

UK) according to the manufacturer’s instructions. As a mea-sure of psychological stress, state anxiety levels wereassessed before as well as 3 and 6 h after LPS/saline injectionby using the German version of the State-Trait-Anxiety Inven-tory (STAI, state version; Laux et al., 1981).

), plasma IL-6 levels (C), salivary a-amylase activity (D), plasmasystolic/diastolic blood pressure (H), and state anxiety (I) before) or placebo (saline) in the same subjects (n = 13) on two differentsults of post hoc computed paired t-tests (*p < 0.05; **p < 0.01;

1822 J.-S. Grigoleit et al.

2.4. Statistical analysis

The Kolmogorov—Smirnov test was used to determinewhether the data meet the assumption of normality. Log-transformations were applied to normalize skewed data.Differences between LPS and placebo conditions were ana-lyzed by two-way repeated measures ANOVA followed by posthoc paired t-tests. Greenhouse—Geisser correction was per-formed when the assumption of sphericity was violated. Onlysignificant interaction effects (time � treatment) arereported. For sAA, NE and cortisol, the area under the totalresponse curve with respect to the ground (AUCG) was cal-culated using the trapezoid formula (Pruessner et al., 2003).Pearson’s correlation coefficient was used to assess correla-tions between physiological measures. Statistical analyseswere performed with PASW Statistics 18 (SPSS, Chicago, IL)and GraphPad Prism 5 (GraphPad Software, San Diego, CA).Results are expressed as mean � SEM and the level of sig-nificance was set at p < 0.05.

3. Results

Administration of bacterial endotoxin to healthy male sub-jects induced an acute systemic inflammatory response asindicated by a transient elevation in body temperature(F[5,60] = 9.70, p < 0.001; Fig. 1A) as well as substantialincreases in WBC counts (F[2.68,32.20] = 46.11, p < 0.001;Fig. 1B) and circulating levels of the pro-inflammatory cyto-kine IL-6 (F[5,60] = 30.69, p < 0.001; Fig. 1C). Salivary a-amylase (sAA) activity significantly increased within 1 h afterLPS administration, reaching peak levels at 2 h, and returningto baseline at 3 h after the injection (F[2.42,28.99] = 7.07,p = 0.002; Fig. 1D). Endotoxin challenge also induced signifi-cant increases in plasma NE (F[2.40,28.74] = 8.16, p = 0.001;Fig. 1E) and salivary cortisol concentrations(F[2.28,27.40] = 19.33, p < 0.001; Fig. 1F), peaking at 2and 3 h, respectively, post injection. Heart rate was signifi-cantly elevated from 1 to 4 h following endotoxin adminis-tration (F[5,60] = 7.63, p < 0.001; Fig. 1G), whereas nosignificant alterations in blood pressure were observed (sys-tolic: F[5,60] = 0.74, p = 0.60; diastolic: F[5,60] = 1.18,p = 0.33; Fig. 1H). Levels of self-reported state anxiety (STAI)were significantly elevated at 3 h after endotoxin injection(F[2,24] = 4.57, p = 0.021; Fig. 1I). No significant correlationsbetween LPS-induced changes (delta scores) in sAA activityand plasma NE levels and salivary cortisol levels, respec-tively, were found. However, when computing the area underthe curve with respect to ground (AUCG) from baseline to 6 hpost injection, a significant positive association betweentotal sAA and plasma NE responses was revealed (r = 0.615,p = 0.025). In contrast, no significant correlation betweentotal sAA and salivary cortisol responses (AUCG) wasobserved.

4. Discussion

Intravenous administration of bacterial endotoxin to healthyhuman subjects induced a systemic inflammatory response asevident from an elevation in body temperature, increases incirculating IL-6 levels and WBC counts, as well as activation

of the two major stress systems, i.e., the sympathetic ner-vous system and the hypothalamus-pituitary-adrenal axis,which is in line with previous reports (Reichenberg et al.,2001; Grigoleit et al., 2011). Here we show that endotoxinchallenge also led to a rapid and transient increase in theactivity of the salivary enzyme a-amylase. This confirms andextends the findings of earlier studies reporting acuteincreases in sAA activity levels in response to a variety ofother physiological and psychological stressors (Nater andRohleder, 2009).

The LPS-induced sAA response was characterized by a fastonset, starting at 1 h after endotoxin administration, and anearly termination relative to the inflammatory response. Incontrast to previous studies using other types of stressors(Chatterton et al., 1996; Nater et al., 2006; Thoma et al.,2012), sAA and plasma NE responses to endotoxin challengeshowed distinct time courses. Although reaching peak levelsat the same time, the increase in sAA activity preceded therise in plasma NE, and sAA levels returned to baseline prior tocirculating NE levels. Given these differences in kineticprofiles, the lack of correlation between delta scores insAA and NE levels is not surprising. Nevertheless, we founda significant positive correlation between endotoxin-inducedsAA and plasma NE responses over the total 6-h observationperiod. This suggests that the observed increase in sAAactivity might be related to sympathetic activation, e.g.,through a rapid nerve-mediated stimulation of the salivaryglands. However, plasma NE has some limitations as a markerof SNS activity as it does not allow discrimination betweencentral or peripheral mechanisms of augmentation (Grassiand Esler, 1999). Thus, further conclusions cannot be drawnwithout assessment of additional SNS markers.

In summary, our study shows for the first time that sAAactivity is enhanced during systemic inflammation. Whetherthis response is driven by an increase in sympathetic activityor more generally reflects inflammation induced changes insympathetic—parasympathetic balance remains to be eluci-dated. Together with extended measurements of sympa-thetic and parasympathetic activity, future studies usingthe human endotoxemia model may help to provide newinsights into the mechanisms underlying stress-associatedchanges in sAA activity.

Role of the funding source

This study was supported by a grant from the GermanResearch Foundation (SCHE341/14-1).

Conflict of interest statement

None declared.

Authors’ contributions

All authors conceived and designed the experiments. JSG,JSK, and HE performed the experiments. RO arranged formedical care and safety. JSG and HE analyzed the data. JSG,MS, and HE wrote the initial draft of the manuscript. Allauthors contributed to and have approved the final manu-script.

Salivary a-amylase response to endotoxin administration in humans 1823

Acknowledgements

We thank all our volunteers for participating in this study. Wethank Christina Banner, Anne Muller, Stephanie Jablonowski,and Dr. Alexander Wegner for excellent technical support.Finally, we would like to thank the division for microbialsafety of the Paul-Ehrlich-Institut in Langen, Germany, forthe safety control of the LPS employed in this study.

References

Bosch, J.A., Veerman, E.C., de Geus, E.J., Proctor, G.B., 2011. a-Amylase as a reliable and convenient measure of sympatheticactivity: don’t start salivating just yet! Psychoneuroendocrinol-ogy 36, 449—453.

Chatterton Jr., R.T., Vogelsong, K.M., Lu, Y.C., Ellman, A.B., Hud-gens, G.A., 1996. Salivary alpha-amylase as a measure of endog-enous adrenergic activity. Clin. Physiol. 16, 433—448.

Engler, H., Doenlen, R., Engler, A., Riether, C., Prager, G., Niemi,M.B., Pacheco-Lopez, G., Krugel, U., Schedlowski, M., 2011.Acute amygdaloid response to systemic inflammation. BrainBehav. Immun. 25, 1384—1392.

Grassi, G., Esler, M., 1999. How to assess sympathetic activity inhumans. J. Hypertens. 17, 719—734.

Grigoleit, J.S., Kullmann, J.S., Wolf, O.T., Hammes, F., Wegner, A.,Jablonowski, S., Engler, H., Gizewski, E., Oberbeck, R., Sche-dlowski, M., 2011. Dose-dependent effects of endotoxin on neu-robehavioral functions in humans. PloS ONE 6, e28330.

Laux, L., Glanzmann, P., Schaffner, P., Spielberger, C.D., 1981. DasState-Trait-Angst-Inventar: Theoretische Grundlagen und Han-dlungsanweisungen. Beltz, Weinheim.

Meisel, C., Schwab, J.M., Prass, K., Meisel, A., Dirnagl, U., 2005.Central nervous system injury-induced immune deficiency syn-drome. Nat. Rev. Neurosci. 6, 775—786.

Nater, U.M., La Marca, R., Florin, L., Moses, A., Langhans, W., Koller,M.M., Ehlert, U., 2006. Stress-induced changes in human salivaryalpha-amylase activity — associations with adrenergic activity.Psychoneuroendocrinology 31, 49—58.

Nater, U.M., Rohleder, N., 2009. Salivary alpha-amylase as anon-invasive biomarker for the sympathetic nervous system:current state of research. Psychoneuroendocrinology 34,486—496.

Proctor, G.B., Carpenter, G.H., 2007. Regulation of salivary glandfunction by autonomic nerves. Auton. Neurosci. 133, 3—18.

Pruessner, J.C., Kirschbaum, C., Meinlschmid, G., Hellhammer,D.H., 2003. Two formulas for computation of the area underthe curve represent measures of total hormone concentrationversus time-dependent change. Psychoneuroendocrinology 28,916—931.

Reichenberg, A., Yirmiya, R., Schuld, A., Kraus, T., Haack, M., Morag,A., Pollmacher, T., 2001. Cytokine-associated emotional andcognitive disturbances in humans. Arch. Gen. Psychiatry 58,445—452.

Scannapieco, F.A., Torres, G., Levine, M.J., 1993. Salivary alpha-amylase: role in dental plaque and caries formation. Crit. Rev.Oral Biol. Med. 4, 301—307.

Thoma, M.V., Kirschbaum, C., Wolf, J.M., Rohleder, N., 2012. Acutestress responses in salivary alpha-amylase predict increases ofplasma norepinephrine. Biol. Psychol. 91, 342—348.

Tracey, K.J., 2002. The inflammatory reflex. Nature 420, 853—859.van Stegeren, A., Rohleder, N., Everaerd, W., Wolf, O.T., 2006.

Salivary alpha amylase as marker for adrenergic activity duringstress: effect of betablockade. Psychoneuroendocrinology 31,137—141.