Manganese Bioconcentration inAquatic Insects: Mn Oxide Coatings,Molting Loss, and Mn(II) ThiolScavengingE L I Z A B E T H K . D I T T M A N A N DD A V I D B . B U C H W A L T E R *

Department of Environmental and Molecular Toxicology,North Carolina State University, Box 7633, Raleigh,North Carolina 27695, United States

Received July 2, 2010. Revised manuscript received October14, 2010. Accepted October 19, 2010.

Streams below mountaintop removal-valley fill coal miningoperations often have elevated Mn concentrations, but it remainsunclear if Mn plays a role in biodiversity reduction. Weexamined various aspects of aqueous Mn interactions withaquatic insects exposed to environmentally relevant Mnconcentrations, revealing complex behavior. First, Mnaccumulation rates varied widely among 9 species. A significantpercentage of total Mn accrued (mean 74%, range 24-95%)was associated with the cuticle, predominantly in the form of Mn-oxides, and to a lesser degree Mn(II). Mn II is also absorbedinto tissues, possibly through calcium transporters. Increasedambient calcium concentrations decreased both adsorbedand absorbed Mn accumulation from solution. Though speciesshowed similar Mn efflux rate constants (0.032-0.072 d-1),the primary mode of Mn loss was through molting. Both adsorbedand absorbed Mn is lost during the molt. Subcellularcompartmentalization studies revealed an overwhelmingtendency for internalized Mn to associate with the heat stablecytosolic protein fraction. After short dissolved Mn exposures,intracellular glutathione and cysteine levels were markedlyreduced relative to controls. These findings suggest that Mnexposureresults in transientphysiologicalstress inaquatic insectswhich is likely relieved, in part, during the molting process.

IntroductionManganese is the 12th most common element in the Earth’scrust, yet it is not commonly studied in freshwater ecosystemsrelative to other transition metals such as zinc, copper,cadmium, and mercury (the 24th, 26th, 72nd, and 73rd mostcommon elements in the earth’s crust, respectively (1)). Muchwork has focused on Mn in sediments; however, Mn in cobblebottomed streams is surprisingly understudied and perhapsoverlooked in situations where it co-occurs with more toxicelements. Manganese is a common constituent of ore andcoal mining and smelting related discharges and otherindustrial effluents. Our interest in Mn stems from itsoccurrence at relatively high concentrations below moun-taintop removal - valley fill (MTM-VF) coal mining operationsin Appalachia (2, 3).

The environmental chemistry of Mn is complex. At neutralto mildly alkaline aerobic conditions, such as those found in

Central Appalachian coal mining areas, Mn is expected tooccur in both Mn(II) and Mn(IV) oxidation states (4). Mn(II)is considered to be the most bioavailable form, particularlywhile occurring as the free ion Mn2+. Mn(IV) occurs asrelatively insoluble oxide precipitates under aerobic condi-tions and can be visible as dark coatings on stream bottomsand on the cuticles of stream insects (4). Importantly, theoxidation of Mn(II) to Mn(IV) can be carried out by severalspecies of bacteria and fungi, which often occurs more rapidlythan abiotic oxidation (5-7).

It is increasingly recognized that ecological conditions instreams below MTM-VFs are typically poor (see review (8)),and field studies show that aquatic insect diversity isdecreased as a function of mining impact, with mayfly generabeing particularly affected (2, 3, 9, 10). Changing hydrology(10-12) and degraded water quality (3) are often thought tobe potential causes. Water quality concerns include elevatedpH (3) and elevated concentrations of dissolved solids (8),selenium (13), and manganese (2). Pond et al. (3) report thataverage dissolved Mn concentrations in mine-impactedstreams were more than 5-fold higher (mean 113.4, range6.5-853 µg Mn L-1) than those in streams unaffected bymining (mean 20.9, range <5-55 µg Mn L-1). Mine effluentdirectly discharged into streams can contain Mn concentra-tions ranging up to 4-5 mg L-1. One field study found anegative correlation between richness of Ephemeroptera,Plecoptera, and Tricoptera (EPT) taxa and Mn concentration(2), though interpretation of these findings is confounded bythe presence of other water chemistry parameters (e.g., totaldissolved solids) that are well outside of typical regionalvalues. It therefore remains unclear whether these elevatedMn concentrations contribute to degraded ecological condi-tions in MTM-VF affected streams.

Aquatic insects dominate stream ecosystems and playimportant roles in ecosystem function as a link between thebase of the food web (e.g., algae, bacteria) and higher orderconsumers (e.g., fish, birds). In general, bioassessment(survey) methods that focus on insect communities cannotdetermine which stressors are most responsible for ecologicalimpairment. Moreover, stream insects are generally not wellrepresented in toxicity databases (14) and may not beadequately represented by standard invertebrate toxicitymodels (e.g., daphnids) that either do not occur in streamsor do not share fundamental physiological traits. As we areunaware of any laboratory studies focused upon aquaticinsects and Mn, this paper represents a first step towardunderstanding Mn bioaccumulation and physiological re-sponses in stream insects and is a starting point for futureinvestigations.

Here we explore Mn bioaccumulation kinetics fromsolution in a variety of field collected aquatic insects. We useMichaelis-Menten type kinetic experiments to examine Mntransport in the caddisfly Hydropsyche betteni and comparedissolved uptake and efflux rates in common SouthernAppalachian stream insects. We assess the influence ofcalcium concentration on Mn accumulation. We furtherquantitatively discriminate between Mn adsorbed to insectcuticles (predominantly as oxides) and Mn incorporated intotissues. The subcellular compartmentalization of Mn iscompared among several species, and the effect of Mnexposure on thiol status is examined. Together, these studiesprovide a framework for understanding the potential effectsof Mn on stream insects.

* Corresponding author phone: (919)513-1129; fax: (919)515-7169;e-mail: david_buchwalter@ncsu.edu.

Environ. Sci. Technol. 2010, 44, 9182–9188

9182 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 23, 2010 10.1021/es1022043 2010 American Chemical SocietyPublished on Web 11/04/2010

Methods and MaterialsInsect Collection and Handling. Insect larvae were fieldcollected using a D-frame kicknet on several dates in 2009and 2010 from two sites in North Carolina (Basin Creek andEno River) and transported back to the lab in a coolercontaining river water and substrate (cobbles). Insects weresorted in the lab, and voucher specimens were preserved inethanol for species verification. Live animals from eachcollection were allowed to acclimate in American Society forTesting and Materials (ASTM) artificial soft water (ASW) (48mg L-1 NaHCO3, 30 mg L-1 CaSO4 ·2H2O, 30 mg L-1 MgSO4,and 2 mg L-1 KCl) for at least twenty-four hours with aerationin a walk-in cold room at 11.5 ( 1.0 °C and a 12:12 h light:dark photoperiod.

Manganese Kinetics in Multiple Species. We compareddissolved Mn uptake kinetics in several common Appalachianspecies - Ephemerella dorothea, Drunella cornutella, Mac-caffertium pudicum, Epeorus vitreus, Acroneuria carolinensis,Pteronarcys sp., and Hexatoma sp. collected from Basin Creek(Wilkes County, NC) in June, 2009. Five individual larvaefrom each species (except Pteronarcys where n ) 3, and D.cornutella where n)10) were used. A solution with a nominalMn concentration of 25.6 µg L-1 in ASW was spiked with54Mn (total activity )0.0632 MBq L-1) as 54MnCl2 (1.4 ng L-1

Mn), with the remainder of Mn added as stable MnCl2. ThepH was adjusted to 7.7 using 0.1 N NaOH. To ensure thatreplicates received identical exposure regimes, bulk solutionswere prepared and distributed to individual 50 mL HDPEbeakers. Each larva was placed into individual beakers with40 mL of solution with aeration and a piece of Teflon meshas substrate. At 2, 4, 6, and 8 h each animal was assayed invivo for radioactivity using a Wallac Wizard gamma counter.After 8 h, each larva was weighed, and those exhibitingsignificant Mn accumulation (E. dorothea, D. cornutella, M.pudicum, A. carolinensis, Pteronarcys sp.) were exposed foran additional 24-48 h to accumulate sufficient radioactivityfor subsequent 10 day depuration experiments (see theSupporting Information). (For molting loss methods andresults see the Supporting Information.)

Dissolved Manganese Uptake in H. betteni. To examinedissolved Mn uptake, Hydropsyche betteni were field collectedfrom the Eno River (Orange County, NC) on January 30 andFebruary 25, 2009. Animals collected in January wereindividually exposed to 10.2, 25.6, 64, 160, and 400 µg Mn L-1

(0.19, 0.47, 1.17, 2.91, and 7.28 µM) in 50 mL HDPE beakerscontaining a total volume of 40 mL ASW with 54Mn radiotracer(0.0646 MBq L-1). February exposures were to a lower rangeof Mn concentrations: 0.26, 0.66, 1.64, 4.10, and 10.2 µg L-1

(4.77, 11.92, 29.82, 74.56, and 186.39 nM) with 54Mn ra-diotracer (0.0608 MBq L-1). Solutions were adjusted to pH7.6-7.7. Animals were exposed as described above. Eachconcentration was represented by five replicates, with eachreplicate consisting of a single larva. After 1, 3, 6, 9, and 24 hof exposure, the animals were removed, rinsed with cleanASW, assayed in vivo for radioactivity, and then returned totheir exposure containers. To ensure that manganese con-centrations remained consistent, exposure water was re-newed after the 9 h time point. After 24 h, the animals wererinsed with clean ASW and weighed. Larvae from the 25.6,64, 160 µg L-1 (0.47, 1.17, and 2.91 µM) exposures were frozenand stored at -20 °C. Insects from the 10.2 and 400 µg L-1

(0.19 µM and 7.28 µM) treatments were individually placedinto 500 mL beakers containing clean ASW for depurationstudies (see the Supporting Information).

Manganese uptake rates were calculated for each indi-vidual at each exposure concentration. The mean of 5replicate slopes was taken as the uptake rate at a givenconcentration and Michaelis-Menten kinetic parameters,Vmax and Km, were calculated using GraphPad Prism 5.0software.

Adsorbed Manganese Oxide Determination. Insects werecollected from the Eno River, NC on January 13, 2010(Cheumatopsyche spp., Maccaffertium modestum, and Iso-nychia spp.) and from Basin Creek on January 28, 2010(Cheumatopsyche spp., Diplectrona modesta, Rhyacophilafuscula, mixed Leptophlebiid species, mixed Maccaffertiumspecies, Acroneuia abnormis, Malirekus hastatus, and Acro-neuria carolinensis). The insects were exposed to 100 µg MnL-1 with 54Mn as a radiotracer as described above (total activity) 0.0629 MBq L-1) and assayed daily for radioactivity. Onday four, they were counted and subjected to an ascorbaterinse (see the Supporting Information). Animals were assayedagain for radioactivity, and percent Mn lost during the rinsewas calculated. Selected species from Basin Creek withsubstantial radioactivity measured after this rinsing proce-dure were used for subcellular fractionation experiments (seebelow).

Calcium Competition. To examine the influence of Ca2+

concentrations on dissolved Mn accumulation in insects,we exposed larvae collected from Basin Creek to 24.2 µg MnL-1 with 54Mn as a radiotracer (total activity ) 0.0637 MBqL-1) under four conditions. ASTM recipes for very soft (VSW),moderately hard (MHW), and very hard (VHW) waters wereprepared in addition to the base very soft water recipe withadditional Ca (VSW+Ca) added (as CaSO4) such that total Cacontent matched the VHW treatment (Table S1). Acroneuriaspp., Ephemerella dorothea, Drunella cornuta, and Maccaf-fertium pudicum were exposed to Mn under each ambientCa condition for 24 h. After 24 h of exposure, animals wererinsed with DI water, assayed for radioactivity, and thenascorbate rinsed to remove Mn oxide precipitates. Insectswere assayed again for radioactivity, weighed, and frozen at-20 °C.

Subcellular Fractionation. Cheumatopsyche spp., Di-plectrona modesta, Maccaffertium spp., Acroneuria abnormis,Acroneuria carolinensis, and Malirekus hastatus from BasinCreek were used to determine the subcellular compartmen-talization of Mn within insect tissue. Single animals werehomogenized in 8 mL of phosphate buffer (pH ) 7.4). Thefractionation scheme was a slightly modified version ofWallace et al. (15) and described elsewhere for insects (16, 17).Fractions obtained were cell debris (including cuticle chitin),organelles, microsomes, heat-denatured proteins (HDP), andheat-stable proteins (HSP) and are reported on the basis ofrecovered Mn.

Thiol Analysis. Hydropsyche betteni were collected fromthe Eno River on September 02, 2009. Animals were exposedto Mn at concentrations of 0 (control), 5, 50, or 500 µg MnL-1 for four days. For each treatment there were threereplicates consisting of 2-3 insects each. After exposure,animals were rinsed with DI water, weighed, and frozen at-20 °C. They were homogenized in 10 mM N-ethylmaleimide(NEM) with 1 µM reserpine in water at a mass (mg) to volume(µL) ratio of 1:19. Samples were centrifuged to remove debris,and the supernatant was analyzed using an LC-MS forconcentrations of reduced and oxidized glutathione andcysteine. Methods were adapted from ref 18 and describedelsewhere (19). Cheumatopsyche spp. and Maccaffertiummodestum collected from the Eno River on January 13, 2010were also used for thiol analysis. Animals were exposed atypical environmental concentration of 100 µg Mn L-1 forfour days and then frozen at -20 °C. For each species, therewere three replicates for both control and Mn exposures,with 2-3 insects per replicate. They were prepared for analysisas described above.

Statistics. Graphs and statistical analysis were completedusing GraphPad Prism (Version 5.0) software. Discrepanciesfrom controls were determined using t tests (R ) 0.05). Vmax

and Km values were determined by using a best-fit Michaelis-

VOL. 44, NO. 23, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 9183

Menten function in Prism. Unless otherwise noted, all valuesare given as mean ( standard deviation.

ResultsManganese Accumulation. Aqueous Mn accumulation ratesranged over 3 orders of magnitude across several speciesexposed to 25.6 µMn L-1. The caddisfly H. betteni and theephemerellid mayfly D. cornutella had the fastest accumula-tion rates of Mn from solution during short-term exposures,780 and 725 ng Mn g-1 h-1, respectively (Figure 1). The tipulidHexatoma sp. and the heptageniid mayfly E. vitreus hadnegligible accumulation from solution. The remaining insectspecies were intermediate in Mn accumulation: anotherephemerellid mayfly E. dorothea and the heptageniid mayflyM. pudicum had accumulation rates of 299 and 64 ng Mn g-1

h-1, respectively (Figure 1). Stonefly species A. carolinensisand Pteronarcys sp. had accumulation rates of 71 and 18 ngMn g-1 h-1, respectively (Figure 1).

Manganese accumulation rates across a wide range ofconcentrations (0.26-400 µg L-1; 4.77 nM-7.28 µM) followeda Michaelis-Menten type kinetics pattern in H. betteni.Estimates of maximal transport rate (Vmax ) 8.58 ( 0.25 µgMn g-1 h-1 (156.1 ( 4.6 nmol Mn g-1 h-1), r2 ) 0.9955) andaffinity (Km ) 266.7( 15.0 µg L-1 (4.86( 0.27 µM Mn)) (mean( standard error) suggest that total Mn accumulation ratesin this species are extremely rapid (Figure 2). This findingsuggests that Mn transport in H. betteni occurs via a lowaffinity, high capacity transport system. We found no evidencefor a complementary high affinity, low capacity transportsystem at concentrations as low as 0.26 µg L-1 (4.769 nM)(Figure 2, inset).

As these initial experiments with H. betteni did notdiscriminate between absorbed (true uptake) and adsorbedMn accumulation, they should be interpreted as total

accumulation and not true transport. Since all larvae usedin these experiments (excluding those used in efflux studiesas described in the Supporting Information) were archived(frozen), we were able to use ascorbate rinses to “correct”for surface adsorption of Mn oxides. On average, 41.9% oftotal radioactivity in H. betteni was removed with this rinsingprocess, representing loss of surface adsorbed metal.

We further compared the relative contribution of surfaceadsorption (as Mn oxides) and internalized (i.e., absorbed)Mn to total Mn body burdens after aqueous exposures to 11species (Figure 3). On average, species (n ) 11) lost 74 (21.7% of their radioactivity after rinsing. M. hastatus lost thelargest percentage of Mn, 95 ( 1.23%, while mixed Lep-tophlebiid species only lost 24 ( 16.4% of their accumulatedMn. Thus, significant Mn associated with insects seems tobe in the form of Mn oxide coatings on the cuticle.

Influence of Water Chemistry on Mn Absorption andAdsorption. Because calcium concentrations can be elevatedin mine-impacted streams, animals were exposed to 24.2 µgMn L-1 under various ionic strengths and Ca concentrations.In general, both Mn absorption and adsorption decreasedwith increasing ionic strength and Ca concentrations (Figure4). For Acroneuria spp., Mn oxide accumulation after 24 hwas significantly decreased from controls (VSW) by 84.4 (p) 0.0026) and 80.1% (p ) 0.0225) in VHW and VSW+Catreatments, respectively (Figure 4a). Post ascorbate rinse,internalized Mn accumulation was significantly decreasedby 53.2 (p ) 0.0127) and 60.3% (p ) 0.0346) in VHW andVSW+Ca treatments, respectively (Figure 4a). Though notstatistically significant, adsorbed Mn on Drunella cornutawas highest in VSW relative to the other treatments (Figure4c), but absorbed Mn burdens were significantly decreasedin VHW and VSW+Ca treatments by 53.1 (p ) 0.0023) and59.9% (p ) 0.0002), respectively (Figure 4c). Mn oxideformation on Maccaffertium pudicum was reduced signifi-cantly by exposure in VHW and VSW+Ca treatments, withburden reductions of 77.2 (p)0.0005) and 76.2% (p)0.0013),respectively (Figure 4b). The same trend was observed forabsorbed Mn in M. pudicum, but the differences were notstatistically significant (Figure 4b). Adsorbed Mn bodyburdens were significantly reduced by 64.7 (p < 0.0001), 72.4(p < 0.0001), and 42.7% (p < 0.0001) in Ephemerellid dorotheaexposed to MHW, VHW, and VSW+Ca treatments, respec-

FIGURE 1. Manganese body burdens (ng Mn g-1 wet weight) asa function of exposure time for insects exposed to 25.6 µg MnL-1. Data points (n ) 5 except Drunella where n ) 10 andPteronarcys where n ) 3) are mean ( standard deviation. Theslope of each line represents the accumulation rate.

FIGURE 2. Michaelis-Menten modeling of Mn uptake rates(mean ( standard deviation; n ) 5) in Hydropsyche betteni asa function of Mn exposure concentration (r2 ) 0.9955). Inset isthe linear portion of the curve, showing data from Mnconcentrations of 0.26 to 10.2 µg L-1 (4.77 to 186.4 nM).

FIGURE 3. Mean ((standard deviation) body burden (µg Mn g-1

wet weight) of adsorbed and absorbed Mn in several speciesexposed to 100 µg/L for 4 days. Bar sets 1-3 represent animalscollected from the Eno River, NC on January 13, 2010.Subsequent bars represent insects collected from Basin Creek,NC on January 28, 2010. (See Table 1 for subcellularcompartmentalization data for absorbed Mn in selected taxa.)

9184 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 23, 2010

tively (Figure 4d). No significant trends were observed forinternalized Mn accumulation in this species (Figure 4d).

Manganese Elimination. Because bioaccumulation ofmetals from solution is a function of both uptake and efflux,we examined Mn loss kinetics in six stream insect species.Manganese elimination was modest in each species tested,with efflux rate constants ranging from 0.032 day-1 in H.betteni to 0.072 day-1 in E. dorothea (Figure S1).

Interestingly, although the loss of Mn was relatively slowin all species, we observed the apparent loss of nearly allaccumulated Mn during the molting process, with the lostMn recovered in the shed exuvia. Discrepancies betweenlarvae and adult tissue Mn body burdens have also beenfound in field collected samples. Analysis of mayflies fromthe Clinch River (CR) and the Little Emory River (LER) inTennessee show that nymphs had mean body burdens of975 and 1215 µg Mn g-1 dry weight, respectively. However,adults collected from the same sites contained means of only3.017 and 2.933 µg Mn g-1 dry weight (Figure S2). Exuviafrom the subimago to imago molt collected from CR andLER showed mean Mn concentrations of 24.01 and 5.472 µg

Mn g-1 dry weight, respectively (Figure S2), but exuvia fromthe larval to subimago molt were unable to be recovered.

Although we found that a significant portion of totalaccumulated Mn was present in the oxide form that couldbe rinsed from the cuticle with a reducing agent, many speciesstill accumulated substantial internalized Mn concentrations(Figure 3). Subcellular fractionation studies revealed that thecytosol was a major sink for Mn bioaccumulation. The meanpercentage of Mn associated with the cytosol across all specieswas 68.6 ( 18.7% and ranged from 39.9 ( 10.1% (inCheumatopsyche spp.) to 90.8 ( 2.6% (in M. modestum).However, within the cytosol, the distribution of Mn betweenHDP and HSP fractions was remarkably consistent acrossspecies and heavily skewed toward the HSP fractions (Table1). Manganese associated with HSP averaged 96.2 ( 1.8% ofthe total cytosolic content in these six species. The HSPfraction is thought to be dominated by small peptides, thiolssuch as glutathione and cysteine, and metallothionein-likeproteins (20).

To test whether Mn exposure influenced thiol status ininsect tissues, we examined reduced glutathione (GSH)

TABLE 1. Percent of Recovered Mn Found in Each of Five Subcellular Fractions (Cell Debris, Organelles, Microsomes,Heat-Denatured Proteins, and Heat-Stable Proteins) for Six Insect Speciesa

species % cell debris % organelles % microsomes% heat-stable

protein% heat-denatured

proteintotal body

burden (µg g-1) % recovered n

Cheumatopsyche spp. 26.6 ( 1.3 24.0 ( 6.7 9.4 ( 4.7 38.4 ( 9.8 1.5 ( 0.3 59.7 ( 12.6 80.2 ( 6.7 2Diplectrona modesta 26.7 ( 5.1 16.5 ( 5.1 6.7 ( 5.9 48.2 ( 5.6 2.0 ( 0.4 43.0 ( 26.0 82.1 ( 6.6 6Maccaffertium spp. 6.1 ( 2.4 1.8 ( 0.8 1.3 ( 0.6 90.1 ( 2.5 0.7 ( 0.5 10.3 ( 7.9 85.3 ( 4.3 6Acroneuria abnormis 29.1 ( 18.5 3.5 ( 2.2 1.9 ( 0.7 62.2 ( 20.7 3.3 ( 2.3 5.6 ( 2.4 90.4 ( 2.0 6Malirekus hastatus 13.3 ( 1.7 3.8 ( 1.1 1.9 ( 0.7 78.7 ( 1.9 2.2 ( 0.4 1.4 ( 0.5 85.7 ( 6.4 3Acroneuria carolinensis 23.4 ( 15.1 3.6 ( 2.0 2.9 ( 1.4 65.7 ( 18.4 4.3 ( 2.8 1.5 ( 0.8 89.8 ( 5.8 6

a Values given as mean ( standard deviation.

FIGURE 4. Mean adsorbed (black bars, left y-axis) and absorbed (white bars, right y-axis) manganese tissue concentrations in fourinsect species after 24 h dissolved exposures (24.2 µg Mn L-1) in varying water chemistry conditions. VSW ) very soft water;VSW+Ca ) very soft water plus calcium; MHW ) moderately hard water; VHW ) very hard water. Tissue concentrations arepresented on a wet weight basis (mean ( standard deviation).

VOL. 44, NO. 23, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 9185

concentrations in H. betteni. Exposure to dissolved Mn for4 days resulted in a 55.4% decrease in GSH concentration (p) 0.0603) at 5 µg Mn L-1 and a 63.2% (p ) 0.0393) and 66.8%(p ) 0.0288) decrease in GSH at 50 and 500 µg Mn L-1,respectively (Figure 5a). Oxidized glutathione (GSSG) con-centrations in this experiment were below quantificationlimits.

Follow up studies revealed a significant effect of Mnexposure on cysteine and cystine concentrations. After fourdays of exposure to 100 µg Mn L-1, cysteine levels weredecreased 64.9% (p ) 0.0674) in Cheumatopsyche spp. and84.4% (p < 0.0001) in M. modestum (Figure 5b). Similarly,cystine concentrations in Cheumatopsyche spp. and M.modestum decreased by 75.6% (p ) 0.0083) and 44.1% (p )0.0205), respectively. Peak areas revealed similar responsesto Mn exposure in GSH concentrations, but problems withstandards did not allow us to rigorously quantify GSH in thisexperiment.

DiscussionManganese is a relatively common, yet poorly studiedelement in freshwater ecosystems. As we were unable to findany Mn experimental work specifically related to aquaticinsects, this paper represents a first step in understandingMn interactions in this important faunal group. Our studiesreveal that Mn interactions with aquatic insects are multi-faceted and complex.

Manganese accumulation from solution ranged 3 ordersof magnitude across several common insect taxa. This findingis not surprising given similar variability observed for differentinsect species with other metals such as Cd (16) and Zn (21).However, it is intriguing that taxa with relatively fast Mnuptake rates (e.g., Hydropsyche, Ephemerella) are also thosewith generally fast transport rates of Cd and Zn (21). Thisfinding suggests that each of these elements may perhaps betransported by a common transport system - possiblyinvolved in Ca transport.

Detailed kinetic experiments with H. betteni revealed alarge capacity to accumulate Mn from solution. The largeVmax and Km suggests a high capacity, low affinity transportsystem. We found no evidence of a second transport systemin play at very low ambient Mn concentrations. However, weonly performed these detailed kinetics studies under artificialsoft water conditions. Ascorbate rinses (but not EDTA alone)suggest that most Mn associated with the cuticle was in theform of Mn oxides. Additionally, the Mn oxide specific dyeLeucoberbelin Blue verified that Mn oxides were not formingin the water or container walls during these experiments butwere obvious on cuticles. This finding leads us to speculate

that differences in oxide production may be related todistinctive microbial communities residing on the cuticle ofvarious insect species. Some species (e.g., Hexatoma sp.) didnot seem to cultivate any surface bound oxides, whereasothers (e.g., M. hastatus) had as much as 94.8% of their totalMn body burdens as surface bound oxides.

Increasing Ca concentrations profoundly decreased bothabsorbed and adsorbed Mn accumulation, which is consistentwith decreased Mn toxicity at increased water hardness levels(22-24). Paired with our finding that Mn is transported viaa high capacity but low affinity system, our results supportformer studies that Mn2+ transport can occur throughchannels meant for other divalent cations, such as Ca2+

(25, 26). Future studies should examine the influence ofdissolved Mn concentrations on Ca ion uptake and regulation.

Manganese efflux rate constants were relatively consistentacross species, but molting was the primary means by whichMn is eliminated from tissues. During efflux experimentslarvae that molted lost the vast majority of their total Mnbody burdens -considerably more than could be removedby ascorbate rinses alone. This finding is in agreement withCid et al. (27), where considerable Mn loss after molting wasdescribed in the European mayfly (Ephoron virgo) in nature.Our own exuvia analyses from Hexegenia sp. corroboratesthe finding that many different aquatic insects shed inter-nalized and surface-bound Mn during the molt process.

Manganese sequestration in the exoskeleton has beenobserved in the calcified body parts of crustaceans, whichalso lose a substantial amount of accumulated Mn duringthe molting process (28). Absorbed manganese can betransferred through the ectoderm to the cuticle, which woulddecrease the internalized metal pools. This partitioning maybe a detoxification mechanism or a means to harden or fortifythe cuticle or certain cuticular tools such as mandibles(29-31). Thus there is some precedent for physiologicalprocesses in place that allow insects to shuttle Mn to thecuticle. Practically, this molting loss makes traditionalbioaccumulation modeling approaches (e.g. refs 32 and 33)insufficient for Mn and also suggests that field-based valuesof Mn tissue concentrations in insects should be interpretedwith caution.

Despite extensive surface binding, several species stillaccumulated an appreciable amount of Mn in tissues, withg90% of internalized Mn found in the HSP fraction of thecytosol. Conventionally, this fraction is thought to containdetoxified metal (15) since molecules such as metallothioneinand glutathione are found here. It may be that small peptides

FIGURE 5. Thiol analysis. a) Concentrations (nmol GSH g-1 wet weight) of reduced glutathione in H. betteni exposed to Mn for fourdays. b) Concentrations (µmol cysteine g-1 wet weight) of cysteine in Cheumatopsyche spp. (black bars) and Maccaffertiummodestum (white bars) exposed to Mn for four days. Values given as mean ( standard deviation.

9186 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 23, 2010

or thiols play a role a transporting Mn to the cuticle duringthe molting process, though we have no direct evidence ofthis.

Relatively short and ecologically relevant Mn exposuresresulted in marked decreases in thiol concentrations. Forexample, we found decreased concentrations of GSH in H.betteni exposed to as low as 50 µg Mn L-1, consistent withprevious studies finding decreased GSH after Mn exposure(34). We also found reduced cysteine concentrations inCheumatopsyche spp. and M. modestum after Mn exposure.These intracellular thiols play important roles in managingthe cell’s redox state. The strong association of Mn with theHSP fraction coupled with the substantial reduction in thefree concentrations of thiols following Mn exposure providessome evidence for Mn acting as a direct thiol scavenger inaquatic insects. Brief dissolved Mn exposures did not affectthe total antioxidant activity of insect homogenates (asmeasured by the Cayman Chemical kit which uses theoxidation of 2,2′-azino-di-[3-ethylbenzthiazoline sulfonateby metmyoglobin). However, previous reports have shownthat Mn exposure decreased total antioxidant status in ratbrain, the effects of which can be lessened by cysteineaddition (35). As we have yet to examine chronic or dietaryMn exposures, the observation that Mn exposure reducesfree thiol concentrations is alarming, especially consideringthat Mn often co-occurs with other metals and contaminantswhich may act as pro-oxidants (e.g. ref 19).

Together these studies provide a first glimpse of Mninteractions with aquatic insects. We show highly variabletransport rates of dissolved Mn among species and highlightthe important process of Mn oxide formation on the cuticlesof different insect species. We provide unequivocal evidenceof Mn loss during the molting process and show a stronginteraction of Mn with heat stable cytosolic proteins andthiols - the latter of which may constitute a significantphysiological stressor. Yet several issues remain unclear andrequire further study. For example, it remains unknownwhether surface Mn oxide coating affects insects’ ability toexchange gases or salts with the surrounding water. Moreover,it is unclear whether dietary Mn (as either Mn(II) or Mn-oxides) is bioavailable or associates with thiols. Thus thereis much to learn about Mn interactions with this importantgroup of ecological indicators.

AcknowledgmentsThe authors appreciate the taxonomic assistance of WilliamCrouch and Eric Fleek, NC DENR. Norm Glassbrook providedthiol analyses. Lingtian Xie prepared exuvia samples foranalysis. Gerald LeBlanc, Justin Conley, Monica Poteat,Kyoung Sun Kim, and Nick Cariello and anonymous reviewersprovided valuable editorial comments. This work was sup-ported by US EPA (83425501-0) and the College of Agricultureand Life Sciences, NCSU.

Supporting Information AvailableDetails of water chemistries, taxonomic information, elimi-nation rates, ascorbate rinses, and molting loss are provided.This information is free of charge via the Internet at http://pubs.acs.org.

Literature Cited(1) CRC Handbook of Chemistry and Physics, 90th ed.; Lide, D. R.,

Eds.; CRC Press/Taylor and Francis: Boca Raton, 2009.(2) Hartman, K. J.; Kaller, M. D.; Howell, J. W.; Sweka, J. A. How

much do valley fills influence headwater streams. Hydrobiologia2005, 532, 91–102.

(3) Pond, G. J.; Passmore, M. E.; Borsuk, F. A.; Reynolds, L.; Rose,C. J. Downstream effects of mountaintop coal mining: compar-ing biological conditions using family- and genus-level mac-roinvertebrate bioassessment tools. J. N. Am. Benthol. Soc. 2008,27 (3), 717–737.

(4) Chiswell, B.; Mokhtar, M. B. The speciation of manganese infreshwaters. Talanta 1986, 33 (8), 669–677.

(5) Hastings, D.; Emerson, S. Oxidation of manganese by spores ofa marine Bacillus - kinetic and thermodynamic considerations.Geochim. Cosmochim. Acta 1986, 50 (8), 1819–1824.

(6) Nealson, K. H.; Tebo, B. M.; Rosson, R. A. Occurrence andmechanisms of microbial oxidation of manganese. Adv. Appl.Microbiol. 1988, 33, 279–318.

(7) Tebo, B. M.; Ghiorse, W. C.; vanWaasbergen, L. G.; Siering, P. L.;Caspi, R. Bacterially mediated mineral formation: Insights intomanganese(II) oxidation from molecular genetic and biochemi-cal studies. Geomicrobiology: Interactions Between Microbes andMinerals 1997, 35, 225–266.

(8) Palmer, M. A.; Bernhardt, E. S.; Schlesinger, W. H.; Eshleman,K. N.; Foufoula-Georgiou, E.; Hendryx, M. S.; Lemly, A. D.; Likens,G. E.; Loucks, O. L.; Power, M. E.; White, P. S.; Wilcock, P. R.Mountaintop mining consequences. Science 2010, 327 (5962),148–149.

(9) Pond, G. J. Patterns of Ephemeroptera taxa loss in Appalachianheadwater streams (Kentucky, USA). Hydrobiologia 2010, 641(1), 185–201.

(10) Simmons, J. A.; Currie, W. S.; Eshleman, K. N.; Kuers, K.;Monteleone, S.; Negley, T. L.; Pohlad, B. R.; Thomas, C. L. Forestto reclaimed mine land use change leads to altered ecosystemstructure and function. Ecol. Appl. 2008, 18 (1), 104–118.

(11) McCormick, B. C.; Eshleman, K. N.; Griffith, J. L.; Townsend,P. A. Detection of flooding responses at the river basin scaleenhanced by land use change. Water Resour. Res. 2009, 45.

(12) Negley, T. L.; Eshleman, K. N. Comparison of stormflowresponses of surface-mined and forested watersheds in theAppalachian Mountains, USA. Hydrol. Process. 2006, 20 (16),3467–3483.

(13) Conley, J. M.; Funk, D. H.; Buchwalter, D. B. Seleniumbioaccumulation and maternal transfer in the mayfly Centrop-tilum triangulifer in a life-cycle, periphyton-biofilm trophicassay. Environ. Sci. Technol. 2009, 43 (20), 7952–7957.

(14) Baird, D. J.; Van den Brink, P. J. Using biological traits to predictspecies sensitivity to toxic substances. Ecotoxicol. Environ. Saf.2006.

(15) Wallace, W. G.; Lee, B. G.; Luoma, S. N. Subcellular compart-mentalization of Cd and Zn in two bivalves. I. Significance ofmetal-sensitive fractions (MSF) and biologically detoxified metal(BDM). Mar. Ecol.: Prog. Ser. 2003, 249, 183–197.

(16) Buchwalter, D. B.; Cain, D. J.; Martin, C. A.; Xie, L.; Luoma, S. N.;Garland, T. Aquatic insect ecophysiological traits reveal phy-logenetically based differences in dissolved cadmium suscep-tibility. Proc. Natl. Acad. Sci. U.S.A. 2008, 105 (24), 8321–8326.

(17) Martin, C. A.; Luoma, S. N.; Cain, D. J.; Buchwalter, D. B.Cadmium ecophysiology in seven stonefly (Plecoptera) species:Delineating sources and estimating susceptibility. Environ. Sci.Technol. 2007, 41 (20), 7171–7177.

(18) Iwasaki, Y.; Hoshi, M.; Ito, R.; Saito, K.; Nakazawa, H. Analysisof glutathione and glutathione disulfide in human saliva usinghydrophilic interaction chromatography with mass spectrom-etry. J. Chromatogr., B 2006, 839 (1-2), 74–79.

(19) Xie, L.; Flippin, J. L.; Deighton, N.; Funk, D. H.; Dickey, D. A.;Buchwalter, D. B. Mercury(II) bioaccumulation and antioxidantphysiology in four aquatic insects. Environ. Sci. Technol. 2009,43 (3), 934–940.

(20) Giguere, A.; Campbell, P. G. C.; Hare, L.; Couture, P. Sub-cellularpartitioning of cadmium, copper, nickel and zinc in indigenousyellow perch (Perca flavescens) sampled along a polymetallicgradient. Aquat. Toxicol. 2006, 77 (2), 178–189.

(21) Buchwalter, D. B.; Luoma, S. N. Differences in dissolvedcadmium and zinc uptake among stream insects: mechanisticexplanations. Environ. Sci. Technol. 2005, 39 (2), 498–504.

(22) Lasier, P. J.; Winger, P. V.; Bogenrieder, K. J. Toxicity ofmanganese to Ceriodaphnia dubia and Hyalella azteca. Arch.Environ. Contam. Toxicol. 2000, 38 (3), 298–304.

(23) Rathore, R. S.; Khangarot, B. S. Effects of water hardness andmetal concentration on a freshwater Tubifex tubifex Muller.Water, Air, Soil Pollut. 2003, 142 (1-4), 341–356.

(24) Stubblefield, W. A.; Brinkman, S. E.; Davies, P. H.; Garrison,T. D. Effects of water hardness on the toxicity of manganese todeveloping brown trout (Salmo trutta). Environ. Toxicol. Chem.1997, 16 (10), 2082–2089.

(25) Crossgrove, J. S.; Yokel, R. A. Manganese distribution across theblood-brain barrier - IV. Evidence for brain influx through store-

VOL. 44, NO. 23, 2010 / ENVIRONMENTAL SCIENCE & TECHNOLOGY 9 9187

operated calcium channels. NeuroToxicology 2005, 26 (3), 297–307.

(26) Fukuda, J.; Kawa, K. Permeation of manganese, cadmium, zinc,and beryllium through calcium channels of an insect musclemembrane. Science 1977, 196 (4287), 309–311.

(27) Cid, N.; Ibanez, C.; Palanques, A.; Prat, N. Patterns of metalbioaccumulation in two filter-feeding macroinvertebrates:Exposure distribution, inter-species differences and variabilityacross developmental stages. Sci. Total Environ. 2010, 408 (14),2795–2806.

(28) Steenkamp, V. E.; Dupreez, H. H.; Schoonbee, H. J.; Vaneeden,P. H. Bioaccumulation of manganese in selected tissues of thefresh-water crab, Potamonautes warreni (Calman), from in-dustrial and mine-polluted fresh-water ecosystems. Hydrobio-logia 1994, 288 (3), 137–150.

(29) Hillerton, J. E.; Robertson, B.; Vincent, J. F. V. The presence ofzinc or manganese as the predominant metal in the mandiblesof adult, stored-product beetles. J. Stored Prod. Res. 1984, 20(3), 133–137.

(30) Quicke, D. L. J.; Wyeth, P.; Fawke, J. D.; Basibuyuk, H. H.; Vincent,J. F. V. Manganese and zinc in the ovipositors and mandibles ofhymenopterous insects. Zool. J. Linn. Soc. 1998, 124 (4), 387–396.

(31) Schofield, R. M. S.; Nesson, M. H.; Richardson, K. A.; Wyeth, P.Zinc is incorporated into cuticular “tools” after ecdysis: Thetime course of the zinc distribution in “tools” and whole bodiesof an ant and a scorpion. J. Insect. Physiol. 2003, 49 (1), 31–44.

(32) Buchwalter, D. B.; Cain, D. J.; Clements, W. H.; Luoma, S. N.Using biodynamic models to reconcile differences betweenlaboratory toxicity tests and field biomonitoring with aquaticinsects. Environ. Sci. Technol. 2007, 41 (13), 4821–4828.

(33) Luoma, S. N.; Rainbow, P. S. Why is metal bioaccumulation sovariable? Biodynamics as a unifying concept. Environ. Sci.Technol. 2005, 39 (7), 1921–1931.

(34) Jang, B. C. Induction of COX-2 in human airway cells bymanganese: Role of PI3K/PKB, p38 MAPK, PKCs, Src, andglutathione depletion. Toxicol. in Vitro 2009, 23 (1), 120–126.

(35) Liapi, C.; Zarros, A.; Galanopoulou, P.; Theocharis, S.; Skandali,N.; Al-Humadi, H.; Anifantaki, F.; Gkrouzman, E.; Mellios, Z.;Tsakiris, S. Effects of short-term exposure to manganese on theadult rat brain antioxidant status and the activities of acetyl-cholinesterase, (Na+, K+)-ATPase and Mg2+-ATPase: Modula-tion by L-cysteine. Basic Clin. Pharmacol. Toxicol. 2008, 103(2), 171–175.

ES1022043

9188 9 ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 44, NO. 23, 2010

Supplemental materials

Table 1. Salt additions for solutions of water with varying hardness. Values are given in mg L-1

of DI water.

Type NaCO3 (mg/L) CaSO4·2H2O (mg/L) MgSO4 (mg/L) KCl (mg/L)

Very soft 12 7.5 7.5 0.5

Very soft + Ca 12 240 7.5 0.5

Moderately hard 96 60 60 4

Very hard 384 240 240 16

Table 2. Taxonomic data and summary information for species used. Mass is given as the mean

± standard deviation of the wet weights in µg.

Order:family Genus species

Experimental

variables

[Mn]

(µg L-1

)

Exposure

Duration

(hours)

Mass

(µg)

Ephemeroptera:Ephemerellidae Ephemerella dorothea Accumulation rates 25.6 8 4.4 ± 0.5

ke 25.6 240 4.4 ± 0.5

Ca competition 24.2 24 4.0 ± 1.3

Drunella cornutella Accumulation rates 25.6 8 6.0 ± 1.5

ke 25.6 240 6.0 ± 1.5

Drunella cornuta Ca competition 24.2 24 4.5 ± 1.4

Ephemeroptera:Heptageniidae Maccaffertium pudicum Accumulation rates 25.6 8 15.4 ± 7.0

ke 25.6 240 15.4 ± 7.0

Ca competition 24.2 24 21.5 ± 10.0

Maccaffertium modestum Adsorbed oxide

quantification

100 96 16.4 ± 3.3

Thiols 100 96 38.9 ± 7.7

Maccaffertium spp. Adsorbed oxide

quantification

100 96 40.0 ± 19.5

Subcellular

fractionation

100 96 40.0 ± 19.5

Epeorus vitreus Accumulation rates 25.6 8 17.8 ± 6.5

Ephemeroptera:Leptophlebiidae Leptophlebiid spp. Adsorbed oxide

quantification

100 96 5.5 ± 3.0

Ephemeroptera:Isonychiidae Isonychia spp. Adsorbed oxide

quantification

100 96 26.8 ± 11.0

Trichoptera:Hydropsychidae Hydropsyche betteni Vmax and Km 0.26 - 400 24 35.8 ± 5.5

Thiols 5 - 500 96 52.6 ± 11.8

Cheumatopsyche spp. Adsorbed oxide

quantification

100 96 6.4 ± 2.3

Subcellular

fractionation

100 96 6.4 ± 2.3

Thiols 100 96 54.8 ± 5.0

Diplectrona modesta Adsorbed oxide

quantification

100 96 19.0 ± 8.2

Subcellular

fractionation

100 96 19.0 ± 8.2

Trichoptera:Rhyacophilidae Rhyacophila fuscula Adsorbed oxide

quantification

100 96 25.2 ± 9.8

Plecoptera:Perlidae Acroneuria carolinensis Accumulation rates 25.6 8 29.1 ± 13.5

ke 25.6 240 29.1 ± 13.5

Adsorbed oxide

quantification

100 96 123.8 ± 78.1

Subcellular

fractionation

100 96 123.8 ± 78.1

Acroneuria abnormis Adsorbed oxide

quantification

100 96 114.1 ± 61.1

Subcellular

fractionation

100 96 114.1 ± 61.1

Acroneuria spp. Ca competition 24.2 24 38.0 ± 43.9

Plecoptera:Perlodidae Malirekus hastatus Adsorbed oxide

quantification

100 96 103.7 ± 17.5

Subcellular

fractionation

100 96 103.7 ± 17.5

Plecoptera:Pteronarcyidae Pteronarcys sp. Accumulation rates 25.6 8 184.6 ± 18.9

ke 25.6 240 184.6 ± 18.9

Diptera:Tipulidae Hexatoma sp. Accumulation rates 25.6 8 97.2 ± 14.2

Methods

Manganese efflux

The acquisition of sufficient radioactivity in larvae exposed to relatively low (0.1864 µM)

and high (7.281 µM) Mn concentrations allowed us to test the premise that efflux rate constants

are independent of tissue concentrations. Following 24 hours of exposure to these concentrations,

H. betteni larvae were added to individual beakers containing 500 mL of clean reconstituted soft

water. The animals and their surrounding water were assayed daily for radioactivity for 10 days.

The water was changed on day four of depuration. The efflux rate constant (ke) was determined

using the equation:

Ct = Ci × e-ket

Where:

Ci = Mn concentration in the animal at time 0 (µg Mn g-1

wet weight)

Ct = Mn concentration in the animal at time t (µg Mn g-1

wet weight)

ke = efflux rate constant (day-1

)

t = time in days

The same procedure was used to assess efflux in six species used for uptake experiments

(Ephemerella dorothea, Acroneuria carolinensis, Drunella cornutella, Maccaffertium pudicum,

Pteronarcys sp). Insects were individually placed into approximately 400 mL of uncontaminated

ASW for depuration as described above, with water changes at 3 day intervals. Animals were fed

periphyton from day 2-7. After ten days, the animals were frozen and stored at -20°C.

Ascorbate Rinse to Remove Mn Oxides

At the end of the exposure period, animals were counted and rinsed thoroughly with

0.1M ascorbate, which is known to reduce metal oxides (such as iron oxide) at much lower

concentrations (1). After the ascorbate rinse, the animals were then rinsed with 0.05M EDTA to

remove the reduced Mn, and were finally washed with deionized water. The use of

Leucoberbelin Blue, which changes color upon contact with Mn oxides, confirmed that this

rinsing process successfully removed Mn oxides from the cuticles (2).

Molting Mn loss from field samples

We were able to obtain Mn tissue concentrations from US Department of Energy (Oak

Ridge) and Tennessee Valley Authority (TVA) biologists from field collected larvae and adult

mayflies (Hexegenia sp.) from two field sites – the Clinch and Emory Rivers in Tennessee.

During the field sampling process, Oak Ridge and TVA scientists archived shed exuvia (sub-

imago to imago molt), which they subsequently sent to us for analysis by ICP-MS.

Ephem

erel

la d

oroth

ea

Dru

nella

corn

utella

Mac

caffer

tium

pudic

um

Hyd

ropsy

che

bette

ni

Acr

oneuria

caro

linen

sis

Ptero

narcy

s sp

.

0.00

0.02

0.04

0.06

0.08

0.10

ke (

day

-1)

Figure 1. Mean (± standard deviation) efflux rate constants, ke (day-1

) for six species of aquatic

insects loaded with Mn and depurated in clean water for ten days.

Nym

ph

Exuvi

a

Imag

o

Nym

ph

Exuvi

a

Imag

o

0

10

20

30

40

50

600

800

1000

1200

1400In

sect

bo

dy b

urd

en

( µµ µµg

Mn

g-1

dry

weig

ht)

Figure 2. Mean (± standard deviation) manganese body burdens (µg Mn g-1

dry weight) of

mayfly nymphs, imagos, and exuvia from the sub-imago to imago molt collected from two sites

in Tennessee. Black bars indicate samples from Clinch River; grey bars represent Little Emory

River.

Literature Cited

(1) Larsen, O. and Postma, D. Kinetics of reductive bulk dissolution of lepidocrocite,

ferrihydrite, and goethite. Geochim. Cosmochim. Acta. 2001, 65 (9), 1367-1379.

(2) Krumbein, W. E. and Altmann, H. J. New Method for Detection and Enumeration of

Manganese Oxidizing and Reducing Microorganisms. Helgol. Wiss. Meeresunters. 1973,

25 (2-3), 347-356.