Trace_elements_in_struvite_equine_enteroliths

Prévia do material em texto

Contents lists available at ScienceDirect
Journal of Trace Elements in Medicine and Biology
journal homepage: www.elsevier.com/locate/jtemb
Analytical methodology
Trace elements in struvite equine enteroliths: Concentration, speciation and
influence of diet
Ashaki A. Rouffa,⁎, George A. Lagerb, Dayana Arruea, John Jaynesc
a Department of Earth and Environmental Sciences, Rutgers University, Newark, NJ 07102, USA
b Department of Geography and Geosciences, University of Louisville, Louisville, KY 40292, USA
c Research Office, University of Louisville, Louisville, KY 40292, USA
A R T I C L E I N F O
Keywords:
Equine
Enterolith
Struvite
Metals
Zinc
XAFS
A B S T R A C T
Equine enteroliths ∼1.5 cm in diameter were collected from an Arabian horse in Louisville, Kentucky, United
States. Scanning electron microscopy (SEM) and light microscope imaging of a sectioned enterolith showed two
distinct regions of concentric growth outward from the central nidus, a small pebble. After initial growth, acidic
colonic fluids permeated the stone inducing recrystallization and alteration of crystals closest to the nidus. A
second growth event, when mineral crystallization was again favorable, produced an outer region of unaltered
crystals at the rim. The mineral was identified as struvite (MgNH4PO4∙6H2O) by powder X-ray diffraction (XRD)
and Fourier transform infrared (FTIR) spectroscopy. Elemental analysis confirmed concentrations of P, Mg and N
consistent with the struvite composition, and detected trace elements Fe (1050–1860 mg kg−1), Mn
(262–280 mg kg−1) and Zn (197–238 mg kg−1). All elements were traced to dietary sources, with the Fe:Mn:Zn
ratio of the enterolith consistent with that of the horse feed. X-ray absorption fine structure (XAFS) spectroscopy
at the Zn K-edge revealed distorted ZnO4 tetrahedra located between crystallographic planes in the struvite
structure forming bidentate linkages to struvite phosphate groups. Emplacement of Zn in structural cavities
likely occurs during struvite crystallization. Trace elements and organic impurities increase susceptibility of the
enterolith to heat-induced decomposition relative to pure struvite, which could be a consideration for treatment.
Results reveal enterolith growth processes, composition and mechanisms of trace metal accumulation that can
inform management and prevention of equine enteroliths.
1. Introduction
Equine enteroliths composed of struvite (MgNH4PO4∙6H2O) are a
major health issue in California and in the Western United States (U.S.).
Removal of large stones ≥12 cm that can obstruct or rupture the colon
represent more than 25% of colic surgeries at University of California
Davis [1]. Reported occurrences of enteroliths in the Eastern United
States are rare, in part because the stones are smaller and usually go
unnoticed in grass pastures. Some of the major risk factors include diet,
for example high magnesium (Mg), nitrogen (N) and phosphorus (P) in
feed, genetic predisposition related to breed, high pH of colonic fluids,
and mineral content of water supply [1]. The enteroliths grow outward
from a central core, or nidus. The nidus is usually a rock fragment, but
can be other material, that acts as a nucleation center for crystal
growth. As the enterolith grows outward, the crystals form concentric
bands similar to those observed in geological mineral concretions. As is
common for minerals crystallized from geological fluids, trace elements
in waste fluids can accumulate in the mineral during crystallization of
biological concretions. In wastewater treatment, controlled struvite
precipitation is used to mitigate the wastewater phosphorus content,
and to recover struvite fertilizer [2,3]. In bench-top experiments with
real and synthetic wastewaters struvite sorbed trace elements, including
zinc (Zn), copper (Cu), arsenic (As) and chromium (Cr), both during
and after precipitation [4–7]. Struvite crystallized from colonic fluids
also sorbs trace elements, with manganese (Mn), iron (Fe), Zn, Cu and
nickel (Ni) detected in equine enteroliths [8,9]. In the current study,
enteroliths from a horse in Louisville Kentucky, Eastern U.S. were col-
lected for investigation. The enterolith mineralogy, crystallization
process, and trace element content were evaluated, and the role of diet
in enterolith formation and composition considered. To better under-
stand trace element sorption to struvite in enteroliths, the speciation
and binding mechanism of Zn was determined. The implications of
trace impurities for thermal stability, and potential heat treatment of
enteroliths was also assessed. Results are of relevance to struvite con-
cretions, including enteroliths and urinary calculi in other organisms
that are formed in the presence of, and accumulate trace elements.
http://dx.doi.org/10.1016/j.jtemb.2017.09.019
Received 28 June 2017; Received in revised form 13 September 2017; Accepted 18 September 2017
⁎ Corresponding author at: Department of Earth and Environmental Sciences, Rutgers University, 101 Warren St Newark, NJ 07102, USA.
E-mail address: ashaki.rouff@rutgers.edu (A.A. Rouff).
Journal of Trace Elements in Medicine and Biology 45 (2018) 23–30
0946-672X/ © 2017 Elsevier GmbH. All rights reserved.
T
http://www.sciencedirect.com/science/journal/0946672X
https://www.elsevier.com/locate/jtemb
http://dx.doi.org/10.1016/j.jtemb.2017.09.019
http://dx.doi.org/10.1016/j.jtemb.2017.09.019
mailto:ashaki.rouff@rutgers.edu
https://doi.org/10.1016/j.jtemb.2017.09.019
http://crossmark.crossref.org/dialog/?doi=10.1016/j.jtemb.2017.09.019&domain=pdf
2. Materials and methods
Enteroliths passed by a senior Arabian gelding in a barn near
Louisville, Kentucky U.S. were collected. The enteroliths were relatively
small with an average diameter of∼1.5 cm and shapes that varied from
spherical to ellipsoidal. Due to the small size, entire enteroliths were
sectioned, and stones minus the nidus were powdered for additional
analysis.
2.1. Petrographic analysis
A whole enterolith was selected and a polished thin section pre-
pared for petrographic analysis. The thin section was analyzed with a
TESCAN VEGA3 scanning electron microscope coupled with energy-
dispersive X-ray spectroscopy (SEM/EDX). The thin section was also
viewed under plane-polarized and cross-polarized light using a Zeiss
polarizing microscope equipped with a digital camera and image-pro-
cessing system.
2.2. Mineralogy
The powdered enterolith, minus the nidus, was analyzed by several
solid characterization techniques to confirm the sample mineralogy and
dominant functional groups. X-ray diffraction (XRD) data were col-
lected from 5 to 60 ° 2-theta (°2θ) in 0.01 ° increments at a counting rate
of 0.3 s per step using a Bruker D8 Advance instrument. Fourier
transform infrared (FTIR) analysis was conducted using a Perkin Elmer
Spectrum 100 instrument and a universal attenuated total reflectance
(ATR) accessory with a ZnSe crystal. Spectra were collected in the
600–4000 cm−1 range, with a resolution of 4 cm−1. The final spectrum
was an average of 20 scans. Results from XRD and ATR-FTIR were
compared to a struvite standard (Alfa Aesar, 98%).
2.3. Simultaneous thermal analysis with evolved gas analysis
The thermal properties of the powdered enterolith were determined
by simultaneous thermal analysis (STA) coupled with FTIR for evolved
gas analysis (EGA). A Netszch Perseus instrument consisting of a STA
449 coupled to a Bruker Alpha FTIR was used with N2(g) as both pro-
tective and sample purge. Solids were heated from 25 to 500 °C at a rate
of 10 °C min−1 and infrared spectra of evolved gases were collected
every 12 s over a range of 600–4000 cm−1 at 8 cm−1 resolution.
Results from STA-EGA were compared to a struvite standard, for which
data were collected under the same conditions.
2.4. Elemental and organic analysis
Enteroliths were submitted to Midwest Laboratories Inc. for com-
plete elemental analysis by inductively coupled plasma atomic emission
spectroscopy (ICP-AES) and nitrogenanalysis using a LECO analyzer.
Two enteroliths from Kentucky were submitted as whole stones minus
nidi. Two samples of California enteroliths, without nidus, from the
inner and outer regions of two stones from Hassel et al. [1] (Fig. 1,
samples 4 and 6) were also analyzed. The samples were acid-digested
by heating in a 6% nitric acid and 1% hydrochloric acid solution.
Blanks were prepared similarly, to ensure that background con-
tamination was not of issue. Certified reference material (CRM) stan-
dards were used for both the calibration and quality control verification
of the instrument. The standards were manufactured according to ISO
9001, ISO 17025, and ISO Guide 34 guidelines and were NIST traceable
to standard reference material (SRM) sources. A laboratory control
sample (LCS) with a recovery tolerance of± 10% was prepared in
parallel with the enterolith samples. The instrument quality control
verification checks were considered valid, and thus reported, only if
within±10% of known values. Duplicates were considered acceptable
when within 10% of the relative percent difference (RPD). The
powdered enterolith used for XRD and FTIR analysis was acid-digested
and analyzed in-house for trace element composition by ICP optical
emission spectroscopy (ICP-OES) using an Agilent 5110 instrument. The
analysis was run in synchronous vertical dual view (SVDV) mode, and
IntelliQuant screening for semi-quantitative determination of elemental
composition conducted prior to the complete analysis. The total organic
carbon (TOC) concentration of this sample was measured by colori-
metric analysis using a TOC direct method reagent test kit and a Hach
DR3900 spectrophotometer.
2.5. Extended X-ray absorption fine structure spectroscopy
Elemental analysis detected Zn as a trace element associated with all
enterolith samples. The binding environment of Zn in the powdered
enterolith sample was determined by extended x-ray absorption fine
structure spectroscopy (EXAFS) analysis. Experiments were conducted
at beamline 12-BM-B at the Advanced Photon Source (APS), Argonne
National Laboratory (ANL), Argonne, IL. The beamline was equipped
with a Si(111) monochromator tuned to the Zn K-edge (9659 eV). Data
were collected in fluorescence mode using a 13 element Ge detector. To
reduce the elastic scattering a Cu filter was placed between the sample
and detector. Data reduction and analysis were performed using Athena
and Artemis software included with the IFEFFIT program [10].
3. Results and discussion
3.1. Enterolith growth patterns
A section through a typical stone shows the central core or nidus, a
chemically altered area surrounding the nidus, and unaltered crystals at
the rim (Fig. 1a). Based on EDX analysis, the nidus is a silicate rock
fragment composed primarily of Mg, aluminum (Al) and silicon (Si). A
thin section through the stone, as viewed with SEM, shows the ring-
shaped alteration area composed of struvite grains within a fine-grained
matrix (Fig. 1b). A magnified view of the upper left portion of the thin
section shows the boundary between the chemically altered region and
crystalline struvite (Fig. 1c). The struvite crystals have well-developed
cleavage, viewed as small square blocks in the images in the unaltered
region. The annular, or ring-shaped area is adjacent to the larger stru-
vite grains, separating the crystalline and altered regions (Fig. 1b, c).
This feature is similar to those found in some natural concretions and
could indicate dewatering—dehydration during crystallization—at the
core of the stone, resulting in volume loss and shrinkage cracks. Acidic
colonic fluids could then permeate the stone dissolving struvite. An-
other growth stage after the stone was sealed and when conditions were
favorable for struvite crystallization might explain the unaltered stru-
vite at the rim.
Images taken using a Zeiss light microscope show additional details
in the chemically altered region of the enterolith (Fig. 1d, e). This re-
gion is permeated by vein-like structures comprised of fine-grained
crystals, with small irregular bright spots in the image representing
individual crystals (Fig. 1d). The veins may have crystallized from
fluids that migrated through cracks in the stone produced by the de-
watering process described above. They consist of P and Mg based on
qualitative chemical analyses (EDX), indicating these may be composed
of struvite. However, the resolution of the SEM/EDX was insufficient to
determine the presence or absence of N (also present in struvite). There
is optically opaque black material in the struvite crystal that parallels
the concentric growth pattern, or fills small radial cracks (Fig. 1d). This
material has not been identified but could represent organic matter
adsorbed on the struvite surface during crystallization.
3.2. Enterolith mineralogy
The XRD analysis of the powdered enterolith, minus the nidus,
shows peaks in the diffraction pattern consistent with a struvite
A.A. Rouff et al. Journal of Trace Elements in Medicine and Biology 45 (2018) 23–30
24
standard, confirming struvite mineralogy (Fig. 2a). The relative in-
tensity of some peaks differs from that of the standard. Increased in-
tensity of peaks, most notably at 21.2 and 33°2θmay be due to effects of
preferred orientation, crystal size and shape effects and/or impurity
atoms associated with these reflections (012, 022). The FTIR spectrum
of the sample also confirms the presence of struvite, with characteristic
vibrational bands for phosphate (ν3 PO43−, 1000 cm−1), ammonium
(ν4 NH4+, 1450 cm−1) and water (760 cm−1, 3500 cm−1) (Fig. 2b)
[4,6,7,11]. Shifting or splitting in the ν3 PO43− band often observed for
incorporation of trace metal impurities such as Zn [6], is not evident for
the sample, indicating no significant impact of any impurities on the
vibrations of this functional group. Compared to the struvite standard,
additional small bands appear between 1250 and 1100 cm−1. These
bands may be attributed to impurities with IR sensitive functional
groups, such as organics, several of which have typical CeO stretching
bands in this region [12].
Fig. 1. a) Section through a typical stone showing the central core (nidus), chemically altered area around nidus consisting primarily of struvite, and unaltered struvite crystals at rim; b)
SEM images of a thin section through the stone showing the annular or ring-shaped alteration area (white arrow), struvite (ST) crystals and nidus, and c) magnified view of the upper left
section of the stone showing boundaries between chemically altered and crystalline struvite (white arrows) and the nidus (N); d) view of the altered area under plane-polarized light using
a Zeiss light microscope showing small crystallized veins (black arrows), and optically opaque black material (green arrows) that parallels the concentric growth pattern, or fills small
radial cracks; e) view using cross-polarized light, showing the fine-grained nature of the veins that permeate the chemically altered area, and small irregular bright spots within the veins
that represent individual crystals. (For interpretation of the references to colour in this figure legend, the reader is referred to the web version of this article.)
Fig. 2. The powdered enterolith compared to a struvite standard a) XRD pattern with degree 2-theta positions and crystallographic planes indicated for select reflections; b) FTIR spectra
with dominant functional groups indicated. c) Select thermal data showing TG curves and percent mass loss, and DSC curves with the temperature of the endothermic peak.
A.A. Rouff et al. Journal of Trace Elements in Medicine and Biology 45 (2018) 23–30
25
3.3. Thermal properties and stability of the enterolith
The enterolith exhibits characteristic thermal properties of struvite.
However, compared to the struvite standard, the absolute thermal
parameters for the enterolith are shifted to lower temperatures
(Table 1, Fig. 2c). Thermogravimetric (TG) analysis shows the en-
terolithbegins to lose mass at a lower temperature (49 °C), with a
smaller total mass loss (45.9%) than the struvite standard (105.7 °C,
53.4%). The peak in the differential of the thermogravimetric curve
(DTG) indicating the highest rate of mass loss, is also shifted to lower
temperature for the enterolith (120.8 °C) compared to the standard
(134.7 °C). The differential scanning calorimetry (DSC) results show a
similar trend, with the endothermic peak and enthalpy of decomposi-
tion at lower values for the enterolith (124.1 °C, 1088 J g−1) relative to
the standard (135.9 °C, 1506 J g−1). Thermal decomposition and as-
sociated mass loss increases absorbance in the FTIR due to the release of
volatiles, ammonia (NH3(g)) and water vapor (H2O(g)). Maximum ab-
sorbance in the FTIR, the peak in the Gram Schmidt (GS) curve, and
temperature of release of both NH3(g) and H2O(g) also occurs at lower
temperature for the enterolith. Impurities associated with the en-
terolith, such as trace metals and organics, may account for more facile
decomposition of the struvite mineral in the sample [4,7]. However,
these additional constituents may in turn reduce the percent volatile
content, and therefore the total mass loss of the sample. Combined,
these results indicate that though the dominant mineralogy of the en-
terolith is struvite, impurities associated with the mineral reduce
thermal stability and increase susceptibility to decomposition and
commensurate release of gases. Sensitivity to temperature, which is
increased by the presence of impurities, may facilitate thermal treat-
ment of enteroliths as has been proposed for struvite urinary calculi
[13].
3.4. Horse history, enterolith composition and role of diet
A senior Arabian gelding (29 years of age) passed the stones in the
stall during feeding. He has very poor dentition and difficulty chewing,
common for horses of this age. Recently, he was diagnosed with pos-
sible PPID (Pituitary Pars Intermedia Dysfunction, or Equine Cushing's
syndrome) based on high ACTH (adrenocorticotropic hormone) levels,
and is currently receiving pergolide. The horse had been recently ex-
periencing colic episodes of unknown origin. Diet consisted of 24/7
pasture, and hay and flax-based horse feed two times daily. Before the
stones were passed, the horse had been fed Purina Hydration hay
blocks, containing at most 40% alfalfa (Purina feeds, pers. comm.). The
owner's veterinarian recommended grass hay rather than Purina hay
blocks based on evidence from enterolith related colic cases in the west
and southwest. In horses with a predisposition to enteroliths (e.g.
Arabians), the approach has been to limit the amount of protein (am-
monia NH3-N is one of the end products of protein metabolism) and Mg.
Both NH3-N and Mg produce a more alkaline environment in the
colon, ideal for struvite crystallization. In addition to the high protein
content, western alfalfa hays are generally high in Mg (for Davis-
Woodland, California area mean Mg% = 0.50 [14]) because of the high
Mg content of some soils. Another possible source of Mg is from the
livestock water, which in the Louisville barn is drawn from a ground-
water well in limestone rock. Analyses of the well water are typical of
karst limestone areas in Kentucky and southern Indiana. The water is
slightly alkaline (pH = 7.5) and considered “very hard” (300 mg L−1
CaCO3). The Mg concentration is 37 mg L−1, which is not unusually
high and significantly lower than that reported for some California
wells, which can exceed 200 mg L−1 [14]. Phosphorous is associated
with grains, such as wheat and flaxseed. The flax-based feed (with beet
pulp and soybean hulls) is balanced for a Ca:P ratio of 2:1, but contains
relatively high amounts of both Mg and P. Results from the diet ana-
lyses in Table 2 indicate that Purina hydration blocks and Standlee
grass hay have essentially the same nutrient profile in terms of crude
Table 1
Thermal parameters derived from simultaneous thermal analysis (STA) coupled with FTIR for evolved gas analysis (EGA) for the powdered enterolith and a struvite standard.
TG DTG DSC FTIR
Onset (°C) Mass loss (%) Peak (°C) Peak (°C) Area (J g−1) Gas: initial detection (°C) GS Peak (°C)
Enterolith 49.0 45.86 120.8 124.1 1088 H2O: 50; NH3: 57 123.2
Struvite 105.7 53.42 134.7 135.9 1506 H2O: 58; NH3: 69 142.7
Table 2
Nutrient analysis of horse feed.a
Purina Hydration Blocks Standlee Timothy Alam (McCauley) Pasture (dry matter, DM) Total Daily Diet (g or mg)b Average Maintenance Body Weight 500 kgc
Crude Protein and Major Minerals in % where %× 10 = g/kg
CP 9⋅7 9⋅9 11 17⋅2 1562 (1333) 630
Ca 0.48 0.52 1.20 0.47 77 (65) 20
P 0.22 0.25 0.61 0.36 45 (39) 14
Mg 0.13 0.17 0.41 0.23 30 (25) 7.5
K 1.45 1.96 n/a 2.08 184 (130) 25
Trace Minerals (mg kg−1)d
Fe 462 315 190 251 3062 (2613) 400
Cu 7 11 30 10 177 (145) 100
Zn 25 40 120 31 646 (528) 400
Mn 51 81 n/a 90 781 (543) 400
Abbreviations: CP (Crude Protein), n/a = not available.
a Nutrient analyses for grass and hay determined by Equi-analytical Laboratory, Ithaca, New York. Alam analysis provided by McCauley Bros., Versailles, Kentucky.
b Total Daily Diet = 11 lbs pasture, 9 lbs Standlee Timothy and 6 lbs Alam, or 11 lbs pasture, 4 lbs Purina hay and 6 lbs Alam, expressed in grams (g) for major minerals and milligrams
(mg) for trace minerals. Diet information from horse owner. Total diet with Purina hydration blocks (no more than 40% alfalfa) in parentheses. Pasture intake estimated from equation: g
DM per kg of body weight = 5.12 sq root (hours grazing)-2.86 = 22.2 g or 11.1 kg DM/500 kg BW for 24 hours grazing time (Paul Siciliano, North Carolina State University). DM amount reduced
to 5 kg DM given the horse’s poor dentition and quality of pasture.
c Nutrient Requirements of Horses, National Research Council, 2007. Dry matter intake 2% body weight, or 10 kg.
d Fe:Cu:Zn:Mn ratio from feed analyses with Purina hydration blocks = 2613:145:528:543 = 18:1:3.6:3.7.
A.A. Rouff et al. Journal of Trace Elements in Medicine and Biology 45 (2018) 23–30
26
protein, Mg and P, suggesting that Purina Hydration hay is not a major
risk factor for enteroliths in this particular case (although not shown,
fiber content for both hays is approximately equal).
Elemental analysis of the enteroliths submitted for ICP-AES is con-
sistent with results from XRD and FTIR, with concentrations of major
elements P, Mg and N essentially those in stoichiometric struvite
(Table 3). These elements are all linked to dietary sources (Table 2).
The presence of trace impurities in the enterolith is also confirmed. The
powdered enterolith contains 0.42% organic carbon, consistent with
the additional features observed in the FTIR spectrum, and which may
be the optically opaque material observed in the polarized-light images
(Fig. 1d). Increased susceptibility to decomposition based on STA-EGA
may also be attributed to organic, as well as trace element impurities.
Detected trace elements include Fe, Mn and Zn (Table 3). This is con-
sistent with the trace elements previously reported in equine enteroliths
[8,9]; though Taylor and Faure [9] did not detect Zn, and the con-
centrations of Fe (25 mg kg−1) and Zn (9 mg kg−1) reported by Blue
and Wittkopp [8] were quite low. The ratio of Fe:Mn:Zn concentrations
in the analyzed Kentucky (KY) enteroliths is 4:1:1 to 9:1:1 (Table 3),
which is close to the 5:1:1 ratio in the diet (Table 2), with some en-
richment of Fe in select enteroliths. The concentration of Fe in diet is
significantly higher than the 40 mg kg−1 feed recommended by the
National Research Council for mature horses [15]. The presence of Fe at
a value of almost seven times the dietary requirement may result in
some excess Fe in waste, and therefore the slightly elevated Fe content
in one of the enteroliths. In human nutrition, over supplementation of
Fe may limit Zn absorption [16]. Although high concentrations of Fe
(Fe overload) have been linked to insulin resistance in horses and other
species [17,18],very little research has focused on the nutrient inter-
action between Fe and Zn in horses. In a study by Lawrence et al. [19],
dietary Fe supplementation (500 and 1000 mg kg−1 feed) lowered Zn
levels in both the serum and liver. In contrast, Pagan [20] concluded
that there was no correlation between Fe content and mineral digest-
ibility when dietary Fe was primarily in the form of Fe oxide. Based on
the similar ratio of elements in the diet and the enterolith, Fe does not
seem to have a significant impact on Zn absorption.
Copper (Cu) is also present in the horse diet, but is not detected in
the enteroliths. High Zn concentrations can interfere with Cu bioa-
vailability. For this reason, Zn:Cu ratios in horse feeds are constrained
and generally vary from 3:1 to 4:1. The Zn:Cu ratio in this horse’s diet is
4:1 (Table 2), therefore Zn is not expected to affect Cu absorption. The
interaction between Fe and Cu has not been studied in the horse but in
human studies based on rodent models Fe can cause Cu deficiency [21].
The Fe:Zn:Cu ratio in diet for Purina hydration blocks is 18:4:1,
whereas the ideal ratio for average maintenance of body weight
(Table 2) is 4:4:1. As noted above, Fe levels are significantly higher than
the NRC requirement. However, if Fe limited Cu bioavailability it is
expected that Cu should be detected in the enterolith, potentially in
excess of the dietary ratio. Assuming, as with the other metals, the ratio
of Cu in the enterolith is consistent with that of the feed, the Cu con-
centration in the enterolith should be approximately 1/3 of the Zn
concentration, ranging from ∼60–70 mg kg−1. This concentration
range is above the analytical detection limit of 20 mg kg−1 for Cu, and
therefore could be quantified if present. That Cu is not detected in the
enterolith suggests that the metal is not present in colonic fluids, pre-
sent at lower than expected concentrations, or not preferentially ac-
cumulated in the enterolith. In a greenhouse wastewater with a dis-
solved Cu:Zn ratio of 1:5, Cu reduced the amount of Zn uptake by
struvite, preferentially binding to struvite phosphate groups and sub-
sequently co-polymerizing with Zn [7]. As Cu has a high affinity for
struvite, and binding is unaffected by the presence of Zn, the con-
centration of Cu present in colonic fluids during enterolith formation is
likely negligible. Therefore, Cu may be absorbed and retained in the
body and not expelled with waste. It is also possible that Cu could have
formed an insoluble compound, or complex, that is present in colonic
fluids but not absorbed and is therefore eliminated as a waste product.
Cu ions released during acid digestion of inorganic salts in the stomach
can form precipitates, such as copper sulfate, in the more alkaline en-
vironment in the small intestine. One source of sulfate is livestock
drinking water but in this case the sulfate concentration is not abnor-
mally high (15 mg kg−1 as SO42−).
The occurrence of Zn (197–238 mg kg−1) in the enteroliths is of
interest, particularly in contrast to Cu, which was not detected.
Analyses of two California enteroliths also revealed the presence of Zn,
though at lower concentrations (Hassel et al. [1], Table 3). Zinc can also
be found associated with human kidney stones, primarily composed of
oxalate and apatite [22], and urinary calculi (struvite) in small rumi-
nants [23]. In the former study, based on a fruit fly model, Zn facilitates
the formation of hydroxyapatite (Ca5(PO4)3(OH)), a possible nidus for
the concretion. Switching off genes that controlled Zn transport in and
out of cells inhibited stone growth in this model, suggesting a genetic
effect. The ruminant study used implanted Zn discs, which accreted
struvite stones. Zinc is known to have an affinity for struvite, forming
adsorbates where Zn is bound to the struvite surface, or becoming in-
corporated into the struvite structure, when present in model solutions
and wastewaters [6,7]. However, the mechanism by which Zn associ-
ates with struvite in enteroliths, and any influence on growth of the
concretion, has not been previously determined.
3.5. Zinc speciation and binding mechanism
To further understand the speciation and binding environment of
Zn, EXAFS data were collected at the Zn K-edge (Fig. 3a,b) for the
powdered sample. The Zn concentration of this sample was measured at
156 mg kg−1 by ICP-OES. Semi-quantitative analysis also detected Fe
and Mn, but no Cu, consistent with the trace element content of the
enteroliths submitted for external analysis (E1-E4, Table 3). Based on
structural parameters derived from the best-fit model, the EXAFS data
can be described by fitting first-shell O and second-shell P atoms
(Table 4). Addition of a Zn shell did not improve the fit, therefore Zn
polymerization is not prevalent. Zinc in the enterolith is in tetrahedral
(4-fold) coordination, with first-shell O atoms located at an average
distance of 1.97 Å. A high sigma squared value (0.014 Å2) denotes
significant variation in the 4 Zn-O bond lengths in this shell, indicating
that the ZnO4 tetrahedron is quite distorted. In Zn-phosphate minerals,
such as liversidgeite and hopeite, Zn is primarily in octahedral (6-fold)
coordination [24,25]. Tetrahedral coordination is typical for Zn oxides/
hydroxides such as zincite and wulfingite [26,27]. This is also a
Table 3
Chemical composition of enteroliths, E1-E4 from horses in Kentucky (KY) and California
(CA). Analyses were carried out by Midwest Laboratories, Inc. using ICP-AES and a LECO
analyzer (nitrogen).
E1 KY E2 KY E3 CA E4 CA
Majors (%)
N 4.59 4.80 5.08 5.04
P 11.80 11.80 11.9 12
Mg 9.51 9.65 9.82 10.2
K 1.06 1.09 1.09 1.43
Ca 0.20 0.27 0.07 0.09
Na 0.01 0.02 0.02 0.02
S n.d. n.d. n.d. n.d.
Trace (mg kg−1)
Fe 1860 1050 447 1580
Mn 280 262 230 160
Zn 197 238 29.7 108
B n.d. 28 24 21
Cu n.d. n.d. n.d. n.d.
KY = Kentucky enteroliths submitted as whole stones minus nidi.
CA = California enteroliths sampled without nidus from inner and outer regions of two
stones from Hassel et al. (2001) Fig. 1 (samples 4 and 6).
n.d. = not detected.
Detection level for zinc and copper = 20 mg kg−1.
A.A. Rouff et al. Journal of Trace Elements in Medicine and Biology 45 (2018) 23–30
27
common configuration for Zn adsorbed to mineral surfaces, including
struvite [6,7]. The data describe a well-ordered second shell of 2–3 P
atoms at a Zn-P distance of 3.40 Å. These results eliminate substitution
of Zn for Mg in the struvite structure, an assumed configuration [9], as
octahedral coordination and a minimum Zn-P distance of ∼4.39 Å are
required for this configuration [28]. The observed distance is the same
as the Zn-P bond for the Zn-2 site in liversidgeite, where Zn forms bi-
dentate complexes with O atoms donated by PO43− groups in the same
plane [25]. Therefore, the Zn-P distance suggests Zn bonded to two
separate PO43− groups (bidentate binuclear) in the same plane. The
closest PO43− groups are at a P-P distance of 6.14 Å along the b axis of
the unit cell (Fig. 3c). Two PO43− groups along this axis can each do-
nate an O atom in the 001 plane to Zn at the observed Zn-P distance.
When complexed with PO43− at the struvite surface, as an adsorbate,
Zn is in tetrahedral coordination with low sigma squared values [7].
Significant distortion of the ZnO4 tetrahedron, due to wide variation in
bond lengths around the average, points to binding at a location in the
mineral structure, likely between crystallographic planes (Fig. 3c).
Because there are large spacings between the crystallographic planes in
struvite, trace metals could easily reside in these locations, as opposed
to substituting for structural ions. The 001 plane has a d-spacing of
11.28 Å, therefore the Zn tetrahedron can easily be accommodated in
the structure between these planes. Zinc atoms in this location would
intersect the 012 and 022 planes for which the associated XRD peaks
show increased intensity relative to the standard (Fig. 2a). The presence
of Zn atoms in the structure maytherefore contribute to the XRD peak
intensities observed for these crystallographic planes.
The configuration of Zn in the enterolith is unique compared to Zn-
struvite configurations previously reported in the literature. For model
wastewaters with Zn present in 0.065–6.54 mg kg−1 (1–100 μM) con-
centrations, both during and after struvite crystallization tetrahedrally
and/or octahedrally coordinated adsorbates and ultimately polymers
were observed [6]. In a greenhouse wastewater with ∼0.65 mg kg−1
Zn, tetrahedrally coordinated Zn polymers forming a Zn-P phase with
Fig. 3. a) The raw chi and b) radial structure EXAFS functions, along with fits to the data obtained at the Zn-K edge for the powdered enterolith sample; c) the proposed configuration of
Zn in the struvite mineral structure (CrystalMaker) based on the EXAFS fit results. The Zn tetrahedron is distorted and may form a bidentate binuclear complex with PO43− groups located
along the b-axis of the unit cell.
Table 4
Fit results to the Zn-K edge EXAFS data for the powdered enterolith sample.
Fit Parameters Shell
O P
N 4 3
R (Å) 1.97 3.40
σ2 (Å2) 0.014 0.002
E0 (eV) 6.81
R factor 0.0197
A.A. Rouff et al. Journal of Trace Elements in Medicine and Biology 45 (2018) 23–30
28
linkages to the struvite structure dominated [7]. In the enterolith Zn
does not polymerize or form a separate Zn-P phase, rather bidentate
complexes that are located in the structure due to the extent of first-
shell O distortion and the well-ordered P shell. Distinct differences in
the speciation and bonding environment of Zn with struvite is con-
tingent upon available metal concentration and crystallization en-
vironment [6]. The Zn concentration in colonic fluids is unknown, but
may well be within the extensive range (0.065-6.54 mg kg−1) explored
in the model wastewater studies. However, the conditions under which
struvite is crystallized in the colon are notably different from the model
wastewaters and collected greenhouse wastewater solutions. In these
latter works, the ions necessary for struvite formation are readily
available in dissolved inorganic form, along with struvite seed material
to induce homogeneous nucleation and rapid precipitation. In colonic
fluids, the nidus, in this case a rock fragment of foreign composition,
acts as a center for heterogeneous nucleation. Struvite crystallization in
the colon is not well understood, but could be facilitated by production
of NH3-N due to microbial activity [8]. For example, the formation of
struvite urinary calculi is linked to bacterial splitting of urea (CH4N2O)
to produce ammoniacal nitrogen (NH3-N), which increases the
NH4+(aq) concentration, pH and formation of PO43−(aq) species
[29,30]. Though urea decomposition is a rapid process, particularly if
catalyzed by urease [29], the rate at which struvite forms under these
conditions is likely to be slower than in the wastewater solutions pre-
viously studied, which are primed with available ions and seed mate-
rial. It is also of note that the heavy metals of interest to this study
inhibit bacterial ureases in the order Cu > Zn > Fe > Mn by
binding to reactive functional groups [31]. If enzyme inhibition is a
factor, this could also indirectly affect the growth rate of struvite.
Therefore, if the role of bacteria in enterolith formation is analogous to
that of urinary calculi, these microbial processes would influence the
growth rate of struvite. However, the mechanisms of alkalinization and
enterolith formation from equine colonic fluids, including the role of
bacteria, are complex and not well understood [8,32]. What is evident
is that the well-defined, large crystals in the enteroliths are indicative of
a slower growth rate and potentially lower saturation [1], when com-
pared to crystals generated from solutions tailored for struvite pre-
cipitation for P recovery [33]. A slower growth process may in turn
facilitate incorporation of Zn into the structure, albeit in between
crystallographic planes, as growth proceeds, rather than relegating the
metal to the mineral surface. Though EXAFS data were only collected
for Zn, similar binding mechanisms may also occur for Mn, which is
present in commensurate concentrations, and potentially Fe.
4. Conclusion
The enterolith mineralogy and trace element content are dictated by
several processes. The type locality for struvite (location at which ori-
ginal material was defined as a mineral species) was a medieval sewer
system with abundant organic matter [34], suggesting that miner-
alization was of bacterial origin. This environment is similar to that in
present day wastewater treatment plants in which struvite formation
becomes favorable when microbial processes release NH3-N to solution.
By analogy with wastewater systems, bacterial metabolism would
produce a very alkaline environment and optimize the composition of
colonic fluids, as dictated by diet, for struvite crystallization. Bacterially
induced formation of struvite urinary calculi is well studied and con-
firmed for other organisms [35]. The same process could occur in the
equine colon if pockets of stagnant fluid around ingested nidi are not
flushed regularly from the intestines. However, factors other than
bacterial fauna may influence both pH regulation and capacity for en-
terolith formation [32], so the exact mechanisms that lead to favorable
conditions for struvite crystallization are unclear. The imaged en-
terolith underwent several distinct growth processes as indicated by the
texture of the struvite crystals. Initial struvite crystallization was fol-
lowed by a stage where acidic colonic fluids permeated the stone, via
the vein-like structures, dissolving struvite and producing the inner
region of altered crystals. A second growth stage after the original en-
terolith was sealed, and when conditions again favored struvite crys-
tallization, would explain the unaltered struvite region near the rim.
During growth, the trace elements Fe, Mn and Zn, present in colonic
fluids accumulated in the enterolith. Zinc is speciated as ZnO4 tetra-
hedra bound to struvite phosphate groups, and located between crys-
tallographic planes in the interstices of the mineral. This is a significant
finding as previous studies of equine enteroliths have suggested sub-
stitution of trace elements for Mg in the struvite structure as the pri-
mary mechanism [9]. Other trace elements such as Mn and Fe, which
are also assumed to substitute for Mg, may also behave similarly to Zn.
The Zn binding configuration may be dictated by the concentration in
colonic fluids and the crystallization process. For the imaged enterolith,
additional experiments are required to determine if Zn (and possibly Fe
and Mn) was emplaced at both stages of growth, shows preferential
accumulation during a single growth stage, and/or was transported in
vein-like structures. Trace impurities, including organics, destabilize
the struvite structure, increasing susceptibility to decomposition. This
may present an opportunity for in-situ heat-induced treatment of the
smaller enteroliths studied here.
However, there are some limitations to the study. In this particular
case, there are so many other factors involved (breed, age, and medical
condition, diet) that it is difficult to draw firm conclusions about the
exact role of Zn and other heavy metals in enterolithiasis. Comparable
metal concentrations for KY and CA enteroliths could indicate similar
growth and mechanisms of metal association for enteroliths from dif-
ferent sources. Case studies of other horses combined with advanced
trace metal analysis of both enteroliths and colonic fluids are required
to confirm this hypothesis. More research is also needed on the role of
bacteria in enterolith formation. This study therefore lays the ground-
work for additional work in this area. In particular, the unique com-
bination of methods and techniques demonstrated here can be used for
further study of metals in enteroliths, and is also transferrableto re-
search on other biological concretions.
Overall, the results of this research provide new insight into the
enterolith growth process, mineralogy, trace element content and the
influence of diet. Results, combined with future work, can be used to
design preventative measures and treatment approaches for affected
horses. The findings are also applicable to other biological concretions
composed of struvite, including zebra enteroliths [36] and urinary
calculi in humans (30% of all calculi), small ruminants, and dogs and
cats (∼75% of all calculi) [37]. Though the current study focuses on
enteroliths from only one horse, tools and techniques used in this study
can provide a basis for further research. A fundamental understanding
of struvite crystallization and trace element sorption is also of relevance
to other systems in which struvite growth may be induced, such as
during wastewater treatment and implementation of P recovery tech-
nologies.
Conflicts of interest
None.
Acknowledgements
This work used the Advanced Photon Source, Argonne National
Laboratory operated under Contract No. DE-AC02-06CH11357. Thanks
to S. Lee and B. Reinhart of Beamline 12BM-B for technical support.
Support was provided by the National Science Foundation Grant Nos.
EAR-1530582 and EAR-1337450. GAL would also like to thank Gayle
Campbell, owner of the Arabian gelding Victor, for donating the en-
terolith samples for this study, Jafar Hadizadeh, University of
Louisville, for discussions of textures in geological concretions, Peter
Schiffman, University of California, Davis, for providing samples from
two California enteroliths for ICP-AES analyses, Thomas Armbruster,
A.A. Rouff et al. Journal of Trace Elements in Medicine and Biology 45 (2018) 23–30
29
University of Bern, Switzerland, for comments on an earlier version of
this paper, and Tatiana Krentsel, Conn Center, University of Louisville,
for SEM/EDX support.
References
[1] D.M. Hassel, P. Schiffman, J.R. Synder, Petrographic and geochemic evaluation of
equine enteroliths, Am. J. Vet. Res. 62 (2001) 350–358.
[2] D.A.C. Manning, Phosphate minerals, environmental pollution and sustainable
agriculture, Elements 4 (2008) 105–108.
[3] K.S. Le Corre, E. Valsami-Jones, P. Hobbs, S.A. Parsons, Phosphorus recovery from
wastewater by struvite crystallization: a review, Crit. Rev. Env. Sci. Technol. 39
(2009) 433–477.
[4] A.A. Rouff, Sorption of chromium with struvite during phosphorus recovery,
Environ. Sci. Technol. 46 (2012) 12493–12501.
[5] N. Ma, A.A. Rouff, The influence of pH and oxidation state on the interaction of
arsenic with struvite during mineral formation, Environ. Sci. Technol. 46 (2012)
8791–8798.
[6] A.A. Rouff, K.M. Juarez, Zinc interaction with struvite during and after mineral
formation, Environ. Sci. Technol. 48 (2014) 6342.
[7] A.A. Rouff, M.V. Ramlogan, A. Rabinovich, Synergistic removal of zinc and copper
in greenhouse waste effluent by struvite, ACS Sustain. Chem. Eng. 4 (2016)
1319–1327.
[8] M.G. Blue, R.W. Wittkopp, Clinical and structural features of equine enteroliths, J.
Am. Veterinary Med. Assoc. 179 (1981) 79–82.
[9] K.S. Taylor, G. Faure, Geochemical study of an equine enterolith, Medina County
Ohio, The Ohio J. Sci. 83 (1983) 54–59.
[10] B. Ravel, M. Newville, ATHENA, ARTEMIS, HEPHAESTUS: data analysis for X-ray
absorption spectroscopy using IFEFFIT, J. Synchrotron Radiat. 12 (2005) 537–541.
[11] E. Banks, R. Chianelli, R. Korenstein, Crystal chemistry of struvite analogs for the
type MgMPO4∙6H2O (M+ = K+, Rb+ Cs+, Tl+, NH4+), Inorg. Chem. 14 (1975)
1634–1639.
[12] R.M. Silverstein, G.C. Bassler, T.C. Morrill, Spectrometric Identification of Organic
Compounds, fourth ed., John Wiley and Sons, New York, 1981.
[13] C.K. Chauhan, M.J. Joshi, In vitro crystallization, characterization and growth-in-
hibition study of urinary type struvite crystals, J. Cryst. Growth 362 (2013)
330–337.
[14] K. Lloyd, H.F. Hintz, J.D. Wheat, H.F. Schryver, Enteroliths in horses, Cornell Vet.
77 (2) (1987) 172–186.
[15] L.M. Lawrence, N.F. Cymbaluk, D.W. Freeman, R.J. Geor, P.M. Graham-Thiers,
A.C. Longland, B.D. Nielsen, P.D. Siciliano, D.R. Topliff, E.V. Valdes, R.J. Van Saun,
Nutrient requirements of horses, Animal Nutrition Series, sixth ed., National
Research Council of the National Academies, Washington, D.C, 2007.
[16] B. Sandstrom, Micronutrient interactions: effects on absorption and bioavailability,
Br. J. Nutr. 85 (Suppl. 2) (2007) S181–S185 (May).
[17] E. Kellon, Iron status of hyperinsulinemic/insulin resistant horses, Proceedings of
the Third European Equine Health and Nutrition Congress, 2006 Mar, Ghent
University. Merelbeke, Belgium, 2006.
[18] B.D. Nielsen, M.M. Vick, P.M. Dennis, A potential link between insulin resistance
and iron overload disorder in browsing rhinoceroses investigated through the use of
an equine model, J. Zoo Wildl. Med. 43 (3 Suppl) (2012) S61–S65 (Sept).
[19] L.A. Lawrence, E.A. Ott, R.L. Asquith, G.J. Miller, Influence of dietary iron on
growth, tissue mineral composition, apparent phosphorous absorption, and che-
mical properties of bone, Proc. Tenth Equine Nutri. Physiol. Soc. Symp. Fort Collins,
CO, 1987 (pp. 563).
[20] J.D. Pagan, Factors affecting mineral digestibility in horses, Proc. Kentucky Equine
Res. Nutr. Conf. Feed Manuf. 7 (1998) 89–104.
[21] L.M. Klevay, Iron overload can induce mild copper deficiency, J. Trace Elem. Med.
Biol. 14 (April (4)) (2001) 237–240.
[22] T. Chi, M.S. Kim, Sven Lang, N. Bose, A. Kahn, L. Flechner, S.D. Blaschko, T. Zee,
G. Muteliefu, N. Bond, M. Kolipinski, S.C. Fakra, N. Mandel, J. Miller,
A. Ramanathan, D.W. Killilea, K. Brückner, P. Kapahi, M.L. Stoller, A Drosophila
model identifies a critical role for zinc in mineralization for kidney stone disease,
PLoS One 10 (5) (2015) e0124150, http://dx.doi.org/10.1371/journal.pone.
0124150G.
[23] M.L. Jones, A.J. Roussel, W.S. Ramsey, Development of a zinc implant-based model
for urolithiasis in goats, Small Rumin. Res. 127 (2015) 97–101.
[24] A. Whitaker, The crystal structure of hopeite, Zn3(PO4)2∙4H2O, Acta Crystallogr. B.
31 (1975) 2026–2035.
[25] P. Elliott, G. Giester, E. Libowitzky, U. Kolitsch, Description and crystal structure of
liversidgeite Zn6(PO4)4∙7H2O, a new mineral from Broken Hill, New South Wales,
Australia, Am. Miner. 95 (2010) 397–404.
[26] R. Stahl, C. Jung, H.D. Lutz, W. Kockelmann, H. Jacobs, Kristallstrukturen und
wasserstoffbruckenbindungen bei beta-Be(OH)2 und epsilon-Zn(OH)2, Z. Anorgani.
Allg. Chem. 624 (1998) 1130–1136.
[27] L. Weber, Die struktur von ZnO, Z. fur Krist. 58 (1923) 398–403.
[28] Ferraris, H. Fuess, W. Joswig, Neutron diffraction study of MgNH4PO4·6H2 (stru-
vite) and survey of water molecules donating short hydrogen bonds, Acta
Crystallogr. B 42 (1986) 253–258.
[29] G. Kramer, H. Klingler, G. Steiner, Role of bacteria in the development of kidney
stones, Curr. Opin. Urol. 10 (2000) 35–38.
[30] J. Prywer, A. Torzewska, Bacterially induced struvite growth from synthetic urine:
experimental and theoretical characterization of crystal morphology, Crys. Growth
Des. 9 (2009) 3538–3543.
[31] B. Krajewska, I. Ureases, Functional, catalytic and kinetic properties: a review, J.
Mol. Catal. B, Enzymatic 59 (2009) 9–21.
[32] D.M. Hassel, S.J. Spier, B.M. Aldridge, M. Watnick, R.A. Argenzio, J.R. Snyder,
Influence of diet and water supply on mineral content and pH within the large
intestine of horses with enterolithiasis, Vet. J. 182 (2009) 44–49.
[33] N. Ma, A.A. Rouff, B.L. Phillips, A 31P NMR and TG/DSC-FTIR investigation of the
influence of initial pH on phosphorus recovery as struvite, ACS Sustain. Chem. Eng.
2 (2014) 816–822.
[34] http://www.mindat.org/loc-1807.html (Accessed 2017).
[35] J. Prywer, M. Olszynski, Bacterially induced formation of infectious urinary stones:
recent developments and future challenges, Curr. Med. Chem. 24 (2017) 292–311.
[36] P. Schiffman, Mineralogical investigation of an enterolith from a Grant’s zebra,
Environ. Geochem. Health.20 (1998) 103–112.
[37] C.K. Chauhan, M.J. Joshi, In vitro crystallization, characterization and growth-in-
hibition study of urinary type struvite crystals, J. Crys. Growth. 362 (2013)
330–337.
A.A. Rouff et al. Journal of Trace Elements in Medicine and Biology 45 (2018) 23–30
30
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0005
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0005
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0010
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0010
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0015
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0015
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0015
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0020
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0020
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0025
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0025
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0025
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0030
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0030
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0035
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0035
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0035
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0040
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0040
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0045
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0045
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0050
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0050
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0055
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0055
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0055
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0060
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0060
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0065
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0065
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0065
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0070
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0070
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0075
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0075
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0075
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0075
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0080
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0080
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0085
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0085
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0085
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0090
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0090
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0090
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0095
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0095
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0095
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0095
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0100
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0100
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0105
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0105
http://dx.doi.org/10.1371/journal.pone.0124150G
http://dx.doi.org/10.1371/journal.pone.0124150G
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0115
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0115
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0120
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0120
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0125
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0125
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0125
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0130
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0130
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0130
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0135
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0140
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0140
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0140
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0145
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0145
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0150
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0150
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0150
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0155
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0155
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0160
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0160
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0160
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0165
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0165
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0165
http://www.mindat.org/loc-1807.html
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0175
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0175
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0180
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0180
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0185
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0185
http://refhub.elsevier.com/S0946-672X(17)30658-2/sbref0185
	Trace elements in struvite equine enteroliths: Concentration, speciation and influence of diet
	Introduction
	Materials and methods
	Petrographic analysis
	Mineralogy
	Simultaneous thermal analysis with evolved gas analysis
	Elemental and organic analysis
	Extended X-ray absorption fine structure spectroscopy
	Results and discussion
	Enterolith growth patterns
	Enterolith mineralogy
	Thermal properties and stability of the enterolith
	Horse history, enterolith composition and role of diet
	Zinc speciation and binding mechanism
	Conclusion
	Conflicts of interest
	Acknowledgements
	References