**3. Results and discussion**

#### **3.1 Dietary paradigm results in a 6-fold difference between** *Artemia***-only and mixed food diets**

Fish fed *Artemia* three times daily consumed an average of 1.2 mg of *Artemia* (51.6 mg wet weight) per feeding for a total of 3.6mg dry mass. Fish reared on the mixed food diet consumed a total of 4.8mg of dry mass per day (1.2mg *Artemia* twice and 2.37mg of Aquatox flakes once). Dietary and environmental iodine levels measured by ICP-MS (Table 1) indicated that the *Artemia* only diet contained just under one-sixth of the total iodine present in the mixed diet (11.48ng and 72.6ng, respectively).


Table 1. Iodide concentrations of study-related salts, feed and supplements. Mean values are listed ± 1 s.d. a(Bidwell and Spotte, 1985), bICP-MS quantified.

The reduction in somatic growth may be the result of diminished endocrine output, possibly caused by an iodine or other nutrient deficiency (Ikeda et al., 1973). The smaller body size was not likely due to underfeeding since the fish were fed three satiating meals of *Artemia*  daily. Further, there was no indication of impaired feeding capacity or precipitous wasting condition in grossly phenotypic fish, despite our expectation that the displacement of the lower jaw bones would negatively affect feeding ability. The *Artemia* fed fish were, in fact, visibly smaller throughout the study.

#### **3.2** *Artemia***-only diet results in smaller, less sexually dimorphic individuals compared with mixed diets**

The body mass and length of fish raised on the *Artemia* diet was significantly reduced compared with fish reared on a mixed diet (Table 2, P<0.0001) despite having made morphometric measurements well after the fish should have normally reached their terminal length (~150 dpf, unpublished data). Sexual dimorphism was clearly evident in the mixed diet population by 4 months of age, but the *Artemia* fed fish were only minimally dimorphic at 10 months of age. Interestingly, a diet of processed flake food alone also

All statistics were performed in SigmaStat XI (Systat Software Inc., San Jose, California). Means were analyzed using Student two-tailed unpaired t-tests or nonparametric rank sum tests as needed. Probability (P) values of <0.05 were considered significant. Incidence levels

**3.1 Dietary paradigm results in a 6-fold difference between** *Artemia***-only and mixed** 

Fish fed *Artemia* three times daily consumed an average of 1.2 mg of *Artemia* (51.6 mg wet weight) per feeding for a total of 3.6mg dry mass. Fish reared on the mixed food diet consumed a total of 4.8mg of dry mass per day (1.2mg *Artemia* twice and 2.37mg of Aquatox flakes once). Dietary and environmental iodine levels measured by ICP-MS (Table 1) indicated that the *Artemia* only diet contained just under one-sixth of the total iodine present

> Crystal Sea Marine Mixa 0.12 Instant Ocean Sea Salta 0.26

System Water (Instant Ocean + salts) 0.73 ± 0.26 SuperSelco Enriched *Artemia*<sup>b</sup> 3.19 ± 1.04

Source Concentration

Seawaterb 0.49-0.58

Aquatox Flakesb 27.38 ± 1.56

Table 1. Iodide concentrations of study-related salts, feed and supplements. Mean values are

The reduction in somatic growth may be the result of diminished endocrine output, possibly caused by an iodine or other nutrient deficiency (Ikeda et al., 1973). The smaller body size was not likely due to underfeeding since the fish were fed three satiating meals of *Artemia*  daily. Further, there was no indication of impaired feeding capacity or precipitous wasting condition in grossly phenotypic fish, despite our expectation that the displacement of the lower jaw bones would negatively affect feeding ability. The *Artemia* fed fish were, in fact,

**3.2** *Artemia***-only diet results in smaller, less sexually dimorphic individuals compared** 

The body mass and length of fish raised on the *Artemia* diet was significantly reduced compared with fish reared on a mixed diet (Table 2, P<0.0001) despite having made morphometric measurements well after the fish should have normally reached their terminal length (~150 dpf, unpublished data). Sexual dimorphism was clearly evident in the mixed diet population by 4 months of age, but the *Artemia* fed fish were only minimally dimorphic at 10 months of age. Interestingly, a diet of processed flake food alone also

(ppm)

were compared using a Yates-corrected Chi-square test.

in the mixed diet (11.48ng and 72.6ng, respectively).

listed ± 1 s.d. a(Bidwell and Spotte, 1985), bICP-MS quantified.

visibly smaller throughout the study.

**with mixed diets** 

**2.6 Statistics** 

**food diets** 

**3. Results and discussion** 


Table 2. Body morphometrics of 160-180 day old fish on study diets. Mean values listed ± 1 s.d.

resulted in reduced terminal lengths when compared to fish fed *Artemia* only or mixed food diets (data not shown).

#### **3.3 Thyroid enlargement leads to jaw bone displacement in fish fed low-iodine diets**

The protruded mandibular jaw phenotype (Fig. 1) was observed within 1 week of initiating the single food *Artemia* diet, with the incidence level increasing logarithmically to 6.3% at 180dpf and reaching a plateau at 10.5% at 10 months of age (n=828 total fish). The incidence level in the population receiving the mixed diet was significantly lower throughout, reaching 3.4% at 10 months (n = 507 total fish, Yates-corrected Chi-square stat: 21.4, α = 0.05, 95% CI: 4.3 – 10.1%). A gradual increase in the extent of mandibular protrusion and degree of localized vascularization was observed throughout the 10 month test period with 78% of the phenotypic fish exhibiting large vascularized growths that protruded an average of 2.1% of their total body length (Fig. 1A). Histologic analyses revealed clear glandular involvement (Fig. 1B) with a distinct rotation of the basihyal jawbone from a sagittal to a more transverse orientation (Fig. 1C). The jaw dysmorphogenesis appeared to be the result of bone displacement by the enlarged thryroid adjacent to the basihyal none rather than a deficiency in osteogenesis, although the normalized Meckel's cartilage and ceratohyal bone lengths were marginally reduced compared to fish fed the mixed diet (Table 2, 19% and 20% reductions for Meckel's and ceratohyal lengths).

Fig. 1. Phenotypic characterization of low iodine diet-reared fish. Gross phenotype (A), histological section stained with hematoxilin and eosin (B), and Alizarin Red bone stain (C). Ventral rotation of the basihyal bone is indicated by an \*.

The presence of an enlarged, hypervascularized tissue mass in proximity to the thyroid gland prompted us to speculate that the overall phenotype was due to an iodine deficiency. Because the teleost thyroid is an unencapsulated colloid-filled follicle (Gundernatsch, 1911), it is often difficult to classify the exact nature of piscine thyroid phenotypes. Nevertheless, several studies in other fish species have shown lower jaw deformities similar to those seen in the *Artemia* fed zebrafish in this study (Gaylord & Marsh, 1912; Marine & Lenhart, 1911; Marine, 1914). The enlarged hypervascularized masses protruding beneath the jaw of New York hatchery salmon were attributed to thyroid carcinomas (Gaylord & Marsh, 1912). Brook trout in a Pennsylvania hatchery also appeared to have simple goiters (Marine, 1914) externally similar in appearance to the phenotype observed in our study. Histochemical analysis of fish reared on our low-iodine diet yielded no indication of cell over proliferation, nuclear atypia, or extra cellular layers typical of carcinomas, but instead showed evidence of pushing borders, organ involvement and hypervascularization typical of simple goiter (Gaylord & Marsh, 1912). Histological analysis also clearly indicated a ventrally-oriented bone displacement associated with the observed lower jaw protrusion, suggesting that the glandular tissue was impinging upon the jawbones and altering their normal orientation. Although our studies terminated after six months, chronic iodine deprivation would have eventually led to 100% incidence of thyroid malformations, however the gradual hypervascularization appeared to exhibit an endpoint at 10.5% of the study fish. Other contributing factors such as environmental stresses and individual fish health may also have modulated the observed phenotypes.

#### **4. Conclusion**

Freshwater fish generally require at least 1 – 4 mg total I kg-1 (Watanabe et al., 1997) and a dietary minimum of 2.8 mg I kg-1 (Lovell, 1979), but plasma iodide levels in freshwater fish vary greatly indicating a wide spectrum in the efficiency of iodine uptake and its use. Given the low 0.1 – 10 µg/l iodine content of freshwater (Eales, 1979) and a metabolic need not met by many natural or artificial environments (Schlumberger, 1954), these fish generally rely more heavily on dietary sources of iodine than do marine fish. Freshwater fish obtain iodine from food by transport through the gut, environmental uptake across the gills, and a very small portion through hormone recycling. Our data suggests that the minimum iodine level required to prevent diminished growth and other apparent iodine-related phenotypes in zebrafish is somewhat lower than for other freshwater fish, falling between the iodine content in our *Artemia* and mixed food diets (being in the range of 0.11 to 0.28 mg I kg-1 per day above the negligible 0.7 ppm environmental exposure). It remains unclear whether iodine supplementation in excess of the mixed diet level (e.g. two *Aquatox* flake feedings per day or *Artemia* supplemented with Kent's iodine) would correct the observed gland enlargement and bone displacement, but previous studies have shown that trout goiters can be attenuated with a potassium iodide supplement in as little as 10 days of exposure time (Marine & Lenhart, 1911; Marine, 1914).

We support the notion that the minimal costs associated with adding iodine-rich flake food to the diet of these fish both is well worth the effort as it results in improved fish health and reduces the time needed to reach sexual maturity.

#### **5. Acknowledgment**

This work was supported by grants from the NIH (#1R01RR023190-01) and AHA (#0555236B) to J.R.H. All experiments were conducted according to guidelines reviewed by the University of Cincinnati Institutional Animal Care and Use Committee under protocol 07-07-30-01.

#### **6. References**

40 Aquaculture

in the *Artemia* fed zebrafish in this study (Gaylord & Marsh, 1912; Marine & Lenhart, 1911; Marine, 1914). The enlarged hypervascularized masses protruding beneath the jaw of New York hatchery salmon were attributed to thyroid carcinomas (Gaylord & Marsh, 1912). Brook trout in a Pennsylvania hatchery also appeared to have simple goiters (Marine, 1914) externally similar in appearance to the phenotype observed in our study. Histochemical analysis of fish reared on our low-iodine diet yielded no indication of cell over proliferation, nuclear atypia, or extra cellular layers typical of carcinomas, but instead showed evidence of pushing borders, organ involvement and hypervascularization typical of simple goiter (Gaylord & Marsh, 1912). Histological analysis also clearly indicated a ventrally-oriented bone displacement associated with the observed lower jaw protrusion, suggesting that the glandular tissue was impinging upon the jawbones and altering their normal orientation. Although our studies terminated after six months, chronic iodine deprivation would have eventually led to 100% incidence of thyroid malformations, however the gradual hypervascularization appeared to exhibit an endpoint at 10.5% of the study fish. Other contributing factors such as environmental stresses and individual fish health may also have

Freshwater fish generally require at least 1 – 4 mg total I kg-1 (Watanabe et al., 1997) and a dietary minimum of 2.8 mg I kg-1 (Lovell, 1979), but plasma iodide levels in freshwater fish vary greatly indicating a wide spectrum in the efficiency of iodine uptake and its use. Given the low 0.1 – 10 µg/l iodine content of freshwater (Eales, 1979) and a metabolic need not met by many natural or artificial environments (Schlumberger, 1954), these fish generally rely more heavily on dietary sources of iodine than do marine fish. Freshwater fish obtain iodine from food by transport through the gut, environmental uptake across the gills, and a very small portion through hormone recycling. Our data suggests that the minimum iodine level required to prevent diminished growth and other apparent iodine-related phenotypes in zebrafish is somewhat lower than for other freshwater fish, falling between the iodine content in our *Artemia* and mixed food diets (being in the range of 0.11 to 0.28 mg I kg-1 per day above the negligible 0.7 ppm environmental exposure). It remains unclear whether iodine supplementation in excess of the mixed diet level (e.g. two *Aquatox* flake feedings per day or *Artemia* supplemented with Kent's iodine) would correct the observed gland enlargement and bone displacement, but previous studies have shown that trout goiters can be attenuated with a potassium iodide supplement in as little as 10 days of exposure time

We support the notion that the minimal costs associated with adding iodine-rich flake food to the diet of these fish both is well worth the effort as it results in improved fish health and

This work was supported by grants from the NIH (#1R01RR023190-01) and AHA (#0555236B) to J.R.H. All experiments were conducted according to guidelines reviewed by the University of Cincinnati Institutional Animal Care and Use Committee under

modulated the observed phenotypes.

(Marine & Lenhart, 1911; Marine, 1914).

**5. Acknowledgment** 

protocol 07-07-30-01.

reduces the time needed to reach sexual maturity.

**4. Conclusion** 

