A comparison of thyroid hormone levels and plasma capillary zone electrophoresis in red‐eared sliders (Trachemys scripta elegans) and map turtles (Graptemys spp.) depending on season and sex

This is an open access article under the terms of the Creative Commons Attribution License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited. © 2020 The Authors. Veterinary Clinical Pathology published by Wiley Periodicals, Inc. on behalf of American Society for Veterinary Clinical Pathology 1Laboklin GmbH & Co. KG, Bad Kissingen, Germany 2Clinic for Small Mammals, Reptiles and Birds, University of Veterinary Medicine Hannover, Hannover, Germany 3Reptile Rescue Center Munich e.V., Munich, Germany 4Department of Veterinary Clinical Sciences, School of Veterinary Medicine, Louisiana State University, Baton Rouge, LA, USA


| INTRODUC TI ON
The thyroid gland is present in all vertebrate species, and the structure and types of hormones produced show minimal variation between the different reptile species. 1 However, the function and physiologic levels of the thyroid hormones have been studied in less detail in reptiles compared with some mammal species. Thyroid hormones play an important role in metabolism, and influence a wide range of physiologic functions, including metabolic rates, growth, regeneration, ecdysis, the production of other endocrine hormones, and reproduction. 1 The thyroid gland is influenced by multiple external factors including, diet, season, temperature, light cycle, and iodine supply. 1,2 At this time, RIs for thyroid hormone levels in chelonians have been calculated for several species of terrestrial tortoises, 3-7 sea turtles, 8 and freshwater turtles. 9 A majority of these studies are limited to tT4 levels and have not been differentiated between the sexes and seasons from which the sample was collected.
Licht et al 9 showed seasonal variations in the tT4 levels of free-living painted turtles in Michigan but did not differentiate between the sexes of the turtles.
Another important facet influencing thyroid hormone levels in the blood is that these hormones bind to several different transport proteins in the blood system. The binding activity and binding capacity of albumin, and the thyroxin-binding protein (TBP) are, however, very different between the various chelonian species. 10 Pavgi et al 11 showed that TBP concentrations were very different between juvenile and adult red-eared sliders (Trachemys scripta elegans), and TBP levels in adult turtles were higher than in juvenile females. In addition, adult females were shown to have significantly higher TBP levels compared with those of adult males. A correlation between TBP concentrations and tT4 binding has been demonstrated. 11 Licht et al 12 showed that distinct changes in tT4 binding lead to changes in plasma T4 concentrations. Similarly, TBP concentrations were shown to be regulated by the concentration of circulating tT4; for instance, lower tT4 levels lead to higher TBP levels. 11 The levels of total protein (TP) and, therefore, the levels of albumin and TBP also varied between individual turtles. Glennemeier et al 13 reported that the percentages of albumin and TBP in the blood of sliders (T scripta) and snapper turtles (Chelydra serpentine) are related to the percentage of bound tT4 and unbound free T4 (fT4), but only the free forms are bioactive. This indicates that the total blood protein and composition of the blood protein fractions, play an important role in thyroid hormone levels.
Thyroid disease has been described in individual case reports. 14 However, the lack of RIs and standardized testing have complicated diagnoses in such cases, highlighting the need for further studies.
Both red-eared sliders and map turtles are native to North America, where they are abundant. These turtles are also commonly kept as pets in North America and other parts of the world. Due to illegal introduction into the wild, these species are also commonly found in freshwater habitats of Europe. To lower the impact on native herpetofauna, trading red-eared sliders has been made illegal in Europe, 15 but many are still kept as pets, which remains legal.
These species, therefore, play an important role in conservation and husbandry.
The goal of this study was to measure plasma protein fractions and thyroid hormone levels for tT4, fT4, tT3, free T3 (fT3), and iodine in captive red-eared sliders (T s elegans) and map turtles (Graptemys spp.) to obtain initial data on the levels and, where possible, to establish RIs, including the evaluation of changes in different seasons. We also aimed to compare the values measured in two different genera of emydid turtles. We hypothesized that the plasma protein fractions and thyroid hormone levels varied between the seasons, sex, and turtle species and that variations in plasma protein and iodine levels influence thyroid hormone levels. It was also hypothesized that there would be a correlation between the levels of the different thyroid hormones.

| Blood collection and sample preparation
Blood samples from red-eared sliders and map turtles were collected from May 2016 to November 2017. The red-eared sliders and map turtles had body weights between 384 and 2867 g and between 170 and 958 g, respectively. Some of the turtles were sampled repeatedly for the study at different times of the year; others were opportunistically included when blood was collected for routine health screening.
Repeat blood collection for this study was carried out under animal welfare permit number TVA-2017-V-18 from the animal welfare committee of the University of Veterinary Medicine, Hannover, accepted on the March 16, 2017. The turtles were kept in two reptile rescue centers in Germany. They were housed in greenhouses with naturalistic indoor ponds, various aquatic plants, and littoral zones with sand for oviposition. The light and temperature conditions were dependent on the natural conditions, but extremes in temperature were mitigated by the greenhouses. The turtles were all fed a varied diet of pellets, small fish, crabs, cuttlebone, various aquatic insects, lettuce, water plants, and wild herbs such as dandelion, field bindweed, and plantains. At the time of blood collection, all turtles were given a general health check (which included an examination of general health, nutritional and care status, the nares, eyes, oral cavity including the mucous membranes, skin, extremities, tail, cloaca, and the form and strength of the shell, as well as a fecal exam) and were all considered clinically healthy. Blood samples were collected from the dorsal coccygeal vein and, in some cases, from the subcarapacial sinus. Lithium heparinized blood and serum samples were collected from red-eared sliders (135 plasma and 127 serum samples) and map turtles (52 plasma and 45 serum samples). Samples with visible lymph dilutions were not included in the study. Blood was transferred into lithium heparinized and serum tubes and stored at 8°C. For thyroid hormone levels and other plasma variables, the samples were centrifuged no later than 4 hours after collection. The plasma was sent over-night to the laboratory and analyzed 1 day after collection. To eliminate samples with lymph contamination and exclude turtles with inflammatory processes from the study, PCVs were measured using microhematocrit capillary tubes that were centrifuged for 5 minutes at 12 000g (Haemofuge, Heraeus Sepatech) and blood smears were microscopically evaluated. The smears were prepared immediately after blood collection, air-dried, and stained using Diff-Quick (Labor + Technik Eberhard Lehmann GmbH). Leukocytes were calculated at 400× magnification in 10 fields. 16 For the differential blood count, 100 cells were evaluated at 1000× magnification.
Samples with leukocyte counts over 19.4 × 10 3 /µL 17 or with PCVs under 0.10 L/L were excluded from the statistical analyses.

| Measurement of thyroid hormone levels, iodine, and plasma protein fractions
The tT4 was measured using an IMMULITE 2000 Xpi Immunoassaysystem canine T4 (KT4) (Siemens Healthcare GmbH) and a cobas 8000 analyzer series module e602 T4 (Roche Diagnostics), the fT4 was measured using a cobas 8000 analyzer series module e602 FT4 II (Roche Diagnostics), and the fT3 was measured using an ADVIA Centaur Immunoassay-system ADVIA Centaur FT3 Assay (Siemens Healthcare GmbH). The samples for tT3 and iodine analyses were stored at −20°C and analyzed 1 month to 1 year after collection. Total T3 was measured using the Triiodothyronine (T3) ELISA Kit (Biomatik Life Science Products and Services). Some of the tT4 and fT4 levels measured in the turtles, especially in samples collected during the summer, were over the detection limit of the tests. The tT4 detection limit of the canine T4 from IMMULITE 2000 Xpi is 15 µg/dL. The samples in which higher concentrations were measured were retested using the tT4 test of the cobas 8000 with a detection limit of 24 µg/ dL. The detection limit of the fT4 test was 100 pmol/L. Samples with higher levels were diluted 1:10 with a physiologic saline solution (0.9% NaCl) and measured again. The test systems used for the measurement of thyroid hormones were validated for the red-eared sliders according to the recommendations of the ASVCP quality control guidelines. Linearity was determined using seven different concentrations of each hormone in turtle plasma, with four replicates measured on each test system. For short-and long-term replication, 21 replicates were measured after a short time (during an 8-hour shift) and after 21 months. For the comparison between the two measurement methods for tT4, 42 turtle plasma measurements were repeated using each test. The tT3 test was appropriate for all animal species according to the manufacturer's specifications and, therefore, only the linearity of this test was determined. Iodine was measured by in-   Tables 1-4, and the RIs for these parameters were calculated using the nonparametric method (10th-90th percentiles). 19 The number of samples was not sufficient for the calculation of the 10th and 90th percentile according to these guidelines at all sample times and, in these cases (<20 samples), only the mean, SD, median, and minimum and maximum values were calculated. For the use of the ANOVA mixed models and t-test, data were normalized by logarithmic transformation. For the calculation of statistical effects, an ANOVA mixed model (SAS) was used, a value of P < .05 was considered the cut-off for significance. The differences between the two detection methods for albumin were calculated using a t-test. A Spearman test (SAS) was used for the calculation of the correlation effects between thyroid hormone levels, iodine, and TP concentrations.

| Validation of test systems
All thyroid hormones (tT4, fT4, tT3, and fT3) showed linear regression within the limits of detection provided by the manufacturer. The short-term replication showed an SD of 0. 25

| Thyroid hormone levels, iodine, and plasma protein analysis in red-eared sliders
A total of 135 samples were collected from red-eared sliders. Four samples were excluded from the statistical analysis due to a low PCV. One hundred and thirty-one samples were evaluated, 34 from males ( Table 1: one in spring, 10 in summer, and 23 in fall) and 97 from females ( Table 2: 10 in spring, 31 in summer, and 56 in fall).
The results for the single male turtle in spring are included, but comparisons based on these values should be considered limited.
The tT4 levels in the males increased from spring to summer and decreased to the lowest levels in fall. In females, the mean tT4 value was highest in the spring and decreased to fall, but the highest individual maximum level was measured in the summer ( Table 2). The fT4 levels in both males and females showed the same pattern as the tT4 levels. In males, the fT4 levels increased from spring to summer and decreased in the fall. In females, the fT4 levels were highest in spring and decreased in the fall, but the highest individual maximum levels were found in the summer. The mean tT4 and fT4 levels in females were higher than in males in spring and summer, but lower in the fall. The tT3 levels increased in males and females from spring to fall. The females had higher mean tT3 levels in summer and fall than the males. The fT3 levels in males increased from spring to summer and decreased in the fall, and, in females, the levels were highest in spring and decreased to fall. Iodine increased in males from spring to summer and decreased in the fall. In female red-eared sliders, the iodine levels were highest in spring and decreased in the fall.
One hundred and thirty-one samples were evaluated by CZE, the results from one female turtle in fall were excluded from further analysis because the electropherogram deviated strongly from those of all other red-eared sliders (Tables 1 and 2). The TP levels decreased in male turtles from spring to summer and increased in the fall. In females, the total protein increased from spring to fall.
Albumin levels detected using the BCG method showed a similar progression; albumin decreased in males from spring to summer and increased in the fall and increased in females from spring to fall. The total albumin levels measured by CZE had a contrary progression in that the levels increased in males continually from spring to fall and decreased in females from spring to summer and increased in the fall. The albumin:globulin (A:G) ratio using CZE increased in males from spring to summer and decreased in the fall. In females, the ratio decreased from spring to summer and increased in the fall. The α-globulin fraction increased in males from spring to summer and decreased slightly in the fall. In females, α-globulin levels developed differently, increasing from spring to summer and decreasing in the fall. In males, β-globulins decreased from spring to summer and increased slightly in the fall. In females, β-globulins increased from spring to summer and decreased in the fall. The γ-globulin fraction decreased slightly in males from spring to summer and increased in the fall. In females, γ-globulins increased continuously from spring to fall.

| Thyroid hormone levels, iodine, and plasma protein analyses in map turtles
Turtles in the genus Graptemys were not further divided into the different species in the present study because of the low number of turtles of different species and the difficulty inherent in differentiating between specific species and hybrids. A total of 52 plasma samples were collected from map turtles, one was excluded from the statistical analysis due to a low PCV. Fifty-one samples were evaluated, 23 from males (Table 3: three in spring, 13 in summer, and seven in fall) and 28 from females (Table 4: three in spring, three in summer, and 22 in fall). Only a lower number of serum samples were available for measuring iodine in male map turtles because lithium heparinized samples were taken first, and the collected blood volume was limited because not enough blood was obtained for all tests in all cases ( Table 3).
The tT4 and fT4 levels in both sexes increased from spring to summer and decreased in the fall, in contrast to the levels in female red-eared sliders. The males had higher tT4 and fT4 levels in spring but lower levels in summer and fall compared with the levels in females. In comparison with the red-eared sliders, the map turtles had lower tT4 and fT4 levels. In males, the tT3 levels increased from spring to fall, like the levels in red-eared sliders.
In females, the tT3 levels decreased slightly from spring to summer and increased in the fall. The fT3 levels in males developed in inverse proportions to the tT3 levels. The fT3 levels in both sexes decreased from spring to fall, like the levels in red-eared sliders.
The iodine levels in males were lowest in spring and increased in the fall. In females, the iodine levels increased from spring to summer and decreased in the fall, similar to the levels in male red-eared sliders.
The results of the protein analyses are based on the same number of samples as the thyroid hormone levels (Tables 3 and 4). The TP levels increased continually in males from spring to fall and increased in females from spring to summer but decreased in the fall.
Albumin measured using the BCG method decreased from spring to summer and increased in the fall. In females, albumin increased in the summer and decreased in the fall. Total albumin measured using CZE increased continually in males and decreased continually in females from spring to fall. The A:G ratio measure using CZE increased in males and decreased in females from spring to fall. The α-globulin fraction decreased in males from spring to summer and increased slightly in the fall. In females, the α-globulin fraction increased continually from spring to fall. In males, the β-and γ-globulins increased in the summer and decreased in the fall. In females, the levels increased continually from spring to fall.
The statistical evaluation of differences between the thyroid hormones and iodine between the species, sexes, and seasons are shown in Tables 5-7.

| Statistical evaluation of the plasma protein fractions and correlation with thyroid hormones
Correlations between protein levels according to species, sex, and season are shown in Table 5. The total albumin levels measured using the BCG and CZE methods showed highly significant differences in male red-eared sliders in the fall (P < .0001) and significant differ-

| D ISCUSS I ON
The results of the present study showed that not all thyroid hormones were correlated with TP and albumin levels in regard to either sex or season. This is in contrast to the results of previous studies, [10][11][12][13]20 in which the binding proteins (TBP and albumin) and tT4 levels were highly correlated. Licht 24,25 There is, however, no specific standard reagent for chelonian albumin available.
Considering these differences and variations, it is most appropriate to measure albumin levels in turtles using electrophoresis, and the BCG test is not recommended in these animals.
Interestingly, CZE led to the differentiation of more peaks in both the sliders and map turtles than described in Hermann's tortoises 23 using the same technology (Figures 1 and 2). In red-eared sliders ( Figure 1) and map turtles (Figure 2), it was possible to differentiate prealbumin, split albumin, split α-globulin, split β-globulin, and single γ-globulin peak. To allow for comparisons with previous studies in red-eared sliders, 26  studied are comparable to those found in Hermann's tortoises using CZE. 23 The main differences between the tortoises and turtles in the present study are the split albumin fractions in most of the turtles examined. A split albumin peak has also been described in green iguanas (Iguana iguana) and could be genetic in origin. 26 Comparisons between the present CZE results and previous studies on emydid turtles show that CZE found split albumin and β-globulin peaks in contrast to agarose gel electrophoresis which was able to differentiate prealbumin, albumin, α1-, α2-, β-, and γ-globulin peaks in free-ranging eastern box turtles (Terrapene carolina carolina). 27 Agarose gel electrophoresis and CAE were both able to differentiate albumin, α-, β-, and γ-globulin peaks in red-eared sliders in May and June. 26 In eastern box turtles, TP and albumin levels were the highest in summer, 27 while the highest albumin peak was found in the fall in red-eared sliders and male map turtles in the present study. The A:G ratio in box turtles did not significantly vary between the seasons, 27 similar to the results obtained in the present study. Significantly higher TP levels in females, as described by Flower et al, 27 were not found in the present study.
Giménez et al 26  In canine and feline medicine, tT4 measurements are often used as an initial screening test for thyroid function. fT4 is used as a more sensitive and specific indicator, and thyroid-stimulating hormone (TSH) is used for further diagnostics. 32 The results of the present study showed that the thyroid hormone levels in turtles are influenced by the season, which is different from most mammals and shows the complexity of these hormones in turtles. In birds and reptiles, tT4 has also been used as an initial screening test. 33,34 Currently, most published studies of thyroid function in chelonians are based on tT4 measurements. 4,[7][8][9] 33 Studies in dogs and cats also show that age influences thyroid hormone levels. 32 A similar effect has also been documented in desert tortoises in July, with higher levels found in adult males compared with those in juveniles and subadult tortoises. 5 All of the turtles included in the present study were adults, and no age differences were documented.
There are numerous factors that can influence thyroid hormone production. Labrada-Martagón et al 8 showed a positive correlation between blood glucose levels and tT4 levels in immature sea turtles (Chelonia mydas), which could be an indication that food intake influences thyroid hormone levels. However, the foods fed to the green sea turtles in that study (shrimp, snails, ctenophores, and seagrasses) 36 are also high in iodine. Increased dietary iodine could, therefore, be responsible for higher iodine blood levels that, in turn, increase thyroid hormone levels. Because iodine is an essential mineral for thyroid hormone synthesis, a deficiency leads to a decrease in types. 38 Germany, in general, is considered an iodine-deficient area, which likely affected the iodine concentrations available to the turtles through the environment and diet, for which a large part consisted of lettuce, wild herbs, and water plants from southern Germany.
However, the iodine concentrations in the turtle blood were relatively high. In the red-eared sliders, in the summer, iodine levels were even higher than those published for American alligators. 37 The present results document the strong variations in the thy-