Accumulation and excretion of manganese ion in the kidney of Mytilus galloprovincialis

The excretion system of mussels consists of kidneys and the pericardium (Martin and Harrison, 1966; Bayne, 1976). Hemolymph is filtered by the auricular wall in the heart (Andrews and Jennings, 1993). Secretion, absorption and storage might be attributed to the kidneys (Martin and Harrison, 1966; Bayne, 1976). In freshwater bivalves, the excretion of hypo-osmotic urine was reported as a means to maintain the osmolality of body fluid. For example, the salt concentration in the final urine is approximately half of the pericardial filtrate (Picken, 1937), and the absorption of chloride and calcium have been detected in Anodonta (Florkin and Duchâteau, 1948). Meanwhile, it is known that marine bivalves are euryhaline (Robertson, 1964). It has been shown that the concentration of electrolytes in the hemolymph of Mytilus were virtually in equilibrium with the external medium (Robertson, 1953). It has been considered that the function of osmoregulation, such as the dilution or concentration of urine, is very limited, and even NH₃/NH₄⁺ was not enriched in Mytilus (Bayne, 1976; Thomsen et al., 2016). Intensive studies have been focused on the accumulation of heavy metal ions such as cadmium, copper, mercury and manganese in bivalves, and the kidney is one of the marked organs that accumulate metals (Bayne, 1976). These studies have been conducted using in vitro techniques such as chemical and histochemical analysis (George et al., 1982), autoradiography (Soto and Cajaraville, 1966), X-ray microanalysis and atomic absorption spectroscopy (Carmichael et al., 1979). As far as we know, no time-resolved in vivo studies have been performed to elucidate the mechanism of the accumulation of heavy metals in the kidney, mainly owing to technical limitations.

In our previous report, we applied a magnetic resonance imaging (MRI) method in a study of the mussel Mytilus galloprovincialis, and also estimated the filtration volume through the auricles of the heart, and determined the direction of the flow of hemolymph in the renopericardial canal (Seo et al., 2014a). The image contrast of MRI is determined mainly by the longitudinal relaxation time (T₁) and the spin-lattice relaxation time (T₂). The T₁ relaxation rate (1/T₁=R₁) is the exponential decay constant of the magnetic moment of the ¹H nuclei of water. The nominal R₁ value for seawater is approximately 0.5 s⁻¹. Some of the heavy metals are paramagnetic, such as the manganese ion (Mn²⁺), which could accelerate T₁ relaxation. Thus, when Mn²⁺ is added to seawater, the R₁ is increased, depending on the concentration. Therefore, R₁ in the kidney should be increased when Mn²⁺ is concentrated in the kidney. In order to test this hypothesis, we (1) measured T₁-weighted MRI (T_1w-MRI) signals and R₁ of the urine in the kidney of mussels enhanced by Mn²⁺, and (2) evaluated the relationship between the Mn²⁺ concentration and R₁ of the urine in the kidney. We also examined the (3) time-dependent R₁ value changes shown by urine in the kidney using three-dimensional T_1w-MRI. The results suggested voluntary excretion of urine from the kidney of mussels.

The Mytilus galloprovincialis Lamarck 1819 used in this study were supplied by Hamasui Co., Ltd (Hiroshima, Japan). These mussels were collected from a subtidal zone and cultivated using a floating suspended culture off the shore of Miyajima, Hiroshima, in April and July 2015. At the laboratory, in two separate 5 liter baths, 10 mussels were kept in each bath for a week in aerated synthetic seawater (salinity 36‰) at room temperature (20–24°C) (Seo et al., 2014b). A total of 15 mussels were used in this MRI study. The length of the mussels was 35.7±0.6 mm (mean±s.e.m). All of the animal experiments in this study were carried out under the rules and regulations of the ‘Guiding Principles for the Care and Use of Animals’ set by the Physiological Society of Japan, and approved by the Animal Research Councils at Dokkyo University School of Medicine.

The MRI examination of the M. galloprovincialis in this study used procedures noted in previous reports (Seo et al., 2014a, 2016). In brief, the mussels were placed in a plastic tube (inner diameter of 22.5 mm), and each mussel was positioned in place using a piece of elastic silicone strip that was inserted at the hinge position of the shell. The mussels were immersed in 15 ml of synthetic seawater without aeration, and the temperature was kept at 20°C. Seawater was exchangeable through another tube set in the bottom of the tube holding the mussel. The ¹H MR images were obtained by a 7 T MRI system (AVANCE III, Bruker Biospin, Ettlingen, Baden-Württemberg, Germany) and equipped with an active shielded gradient (micro2.5) and a 25-mm ¹H birdcage radiofrequency coil.

The T₁ values of the MnCl₂ solutions (1.2–6 mmol l⁻¹) were measured by inversion recovery pulse sequences with a 10-mm ¹H coil at 22°C. The relaxivity of Mn²⁺ was calculated from the slope of a linear regression line [6.3 l (s mmol)⁻¹].

All of the mussels were fixed with 4% paraformaldehyde (PFA) for the histological examinations, and embedded in paraffin wax after dehydration. The paraffin sections were prepared using a slice thickness of 10 µm. The sections were stained with hematoxylin & eosin (H&E). Images were captured and combined using a microscope (BZ-9000, Keyence, Osaka, Japan) with an image-stitching mode.

First, we reviewed the anatomy of M. galloprovincialis using MRI images and light microscopic analysis. The orientation of the organs are shown in a transverse MRI image obtained 3 mm posterior from the auriculoventricular valve (Fig. 1A). The kidneys were located on the dorsal side of the gill, on the inner side of the anterior oblique vein, and on the ventral side of the posterior retractor muscles. The shape of the kidney wall was lobular, and some parts were adjacent to the venous wall with no interstitial space (Fig. 1A–C). In order for a mussel to excrete urine, it is thought that the excretory pore opens to the upper mantle cavity connecting to the exhalant siphon (Fig. 1B). The excretory pore consisted of a short duct surrounded by muscular tissue (Fig. 1C). The kidneys were connected with the pericardium through the renopericardial canal. The renopericardial canal and the anterior oblique vein run side by side (Fig. 1D). Some parts of the wall of the renopericardial canal were lobular and parallel with the anterior oblique vein with no interstitial space (Fig. 1E). The structure of the kidney, heart and adjacent vessels were illustrated in a schematic diagram (Fig. 1F).

Next, we examined the distribution of a heavy metal, manganese ion, in the mussels using MRI. In the mussels fixed with PFA, the anatomical structure of the kidney was detected in the T_1w-MR images (Fig. 2A), but in the living mussels, it was difficult to identify the kidney because T₁ relaxation rate (R₁) values of kidney (0.551±0.035 s⁻¹; mean±s.e.m, n=28) were similar to those of surrounding tissues and seawater (Table 1, Fig. 2B). When a live mussel was immersed in 300 ml aerated seawater containing 50 µmol l⁻¹ MnCl₂, the kidneys and digestive organs were depicted at a higher signal intensity (Fig. 2C). The R₁ of the kidney transiently increased for 2–10 h, then decreased at 24 h of exposure to Mn²⁺ (1.83±0.132 s⁻¹, n=14; Table 1). In a separate experiment, R₁ of the kidney was maintained at almost the same level after 50 h of exposure to Mn²⁺ (1.81±0.301 s⁻¹, n=4). When the mussel was returned to normal seawater, the R₁ of the kidney decreased slowly, and returned to the control level at 72 h (Table 1). Therefore, M. galloprovincialis seemed to concentrate Mn²⁺ in the kidney from the surrounding seawater, and seemed to excrete Mn²⁺ from the kidney.

Considering these results, the T_1w-MR image of the kidney seemed to be contrasted by the accumulated manganese ion. However, it is known that T_1w-MR image intensity is affected by flow, such as the circulation of body fluids (Bock et al., 2001; Seo et al., 2014b). In order to eliminate the effect of flow in the kidney, the mussel was anesthetized using 4% MgCl₂. In a transverse image of the region around the heart, the kidneys and gastrointestinal tract were depicted at a higher signal intensity compared with the PFA fixation, but without Mn²⁺, similar to Fig. 2A (Fig. 3A). After the application of 50 µmol l⁻¹ Mn²⁺, the pericardium, ventricle and auricles of the heart were not enhanced, but the kidneys were clearly contrasted (Fig. 3B). These results indicated the T_1w-MRI signal intensity in the kidneys was not affected by the flow. The 3D reconstructed image of the kidneys showed the whole structure of the kidneys in the same condition as shown in Fig. 3B (Fig. 3C; Movie 1). The anterior–posterior lengths of the right and left kidneys were 16.6 mm and 15 mm, respectively, and the volumes of the right and left kidneys were 24 µl and 14 µl, respectively. It is interesting to note that the pericardial canals were depicted from the middle of the canal to the kidneys, leading to the speculation that the renopericardial canals might participate in the concentration of manganese ion (* in Figs 3C and 1F).

In order to analyze the initial increasing phase of the [Mn²⁺] in the kidney, the accumulation of Mn²⁺ in the kidneys was measured using 24 kidneys of 12 mussels. The T_1w-MR image intensity of the kidneys was continuously/linearly increased for the first 2 h in 12 kidneys of eight mussels. A typical result is shown in Fig. 5A and Movie 2-1. When the normal seawater was replaced by seawater containing Mn²⁺, the signal intensity of the stomach increased instantaneously, which was then followed by the increase in the kidneys after a delay of a few minutes. The rate of increase of the signal intensity was constant, and the rates in the respective kidneys were similar to each other. As shown in the sagittal image of the left kidney (Movie 3), the signal intensity increased in the same pattern as seen in the long axis of the kidney. The image intensity of the seawater increased a bit owing to the 10 µmol l⁻¹ Mn²⁺, then decreased gradually owing to uptake of Mn²⁺ into the mussel. The MR signals in the soft tissues were constant, except for the digestive organs, such as the stomach and the intestine. Because the increase of the T_1w-MR image intensity with a short echo time was proportional to the Mn²⁺ concentration (Fig. S1), the [Mn²⁺] in the kidneys was estimated by the acquired MR image intensity. The means±s.e.m for 12 kidneys are shown in Fig. 5D. The initial increase rate of [Mn²⁺] in the kidney (4.83±0.58 µmol l⁻¹ min⁻¹, n=12) was calculated from the slope of the regression line of [Mn²⁺] at 10–40 min. A linear regression line of the increase in [Mn²⁺] was determined by the increase in [Mn²⁺] at 10–40 min, and the intercept to time axis was defined as the initial delay of [Mn²⁺] increase in the kidney (8.16±0.44 min, n=12).

In contrast, in some kidneys, the maximal image intensity of the kidney fell to below 50 µmol l⁻¹ and increased again after a short interval. A single intensity fall was observed in six kidneys of five mussels, and one example is shown in Fig. 6 and Movie 2-2. In this experiment, the mussel was immersed in 10 µmol l⁻¹ Mn²⁺ seawater for the first 2 h, and thereafter, the 10 µmol l⁻¹ Mn²⁺ seawater was replaced by 20 µmol l⁻¹ Mn²⁺ seawater. At first, [Mn²⁺] in both kidneys increased in the same kinetic pattern, but the [Mn²⁺] of the right kidney dropped to almost the basal level at 58 min. Then, the [Mn²⁺] of the right kidney started to increase again, and [Mn²⁺] increased at a rate similar to that seen in the left kidney. The [Mn²⁺] in the right and left kidneys at 120 min was estimated at 200 and 250 µmol l⁻¹, respectively. When the seawater was replaced by 20 µmol l⁻¹ Mn²⁺ seawater, the signal intensities in both kidneys dropped significantly. Thereafter, the image intensity increased again instantaneously. The average of the three drops was 173.5±26.6 µmol l⁻¹. The fraction of residual urine in the kidney could be estimated as a ratio of Mn²⁺ concentration after and before the drop (13.3±0.3%).

The rest of the six kidneys of three mussels demonstrated transient falls in [Mn²⁺] more than two times. Twelve falls showed drops of more than 75 µmol l⁻¹ (106.2±7.1 µmol l⁻¹), and the average fraction of residual urine was 34.5±5.6%. Typical results are shown in Fig. 7 and Movie 2-3. The right kidney demonstrated transient falls in [Mn²⁺] of the kidney three times, and the left kidney demonstrated transient falls four times. The falls in [Mn²⁺] were not synchronized between the right and left kidneys. However, the falls in [Mn²⁺] were synchronized between four slices from the posterior side to the anterior side of the kidney.

This is the first study to visualize manganese ion accumulation in the urine of kidneys of bivalves using signal enhancement of T_1w-MRI. Because the image intensity and R₁ of the stomach and digestive glands also increased, the Mn²⁺ might have been taken from the gastrointestinal tract, and then filtrated into the pericardium by auricles. There were no signal enhancements in the auricles or the pericardium. Therefore, the Mn²⁺ was filtrated by the auricular wall, and was not concentrated in the auricles or the pericardium. Accumulation of the Mn²⁺ was detected not only in the kidneys, but also in the distal part of the renopericardial canals (* in Fig. 3C). Consider the following evidence. (1) The walls of the renopericardial canal were attached to the anterior oblique vein, and the wall was a convex surface into the anterior oblique vein (Fig. 1E). This is the same in the wall of auricles that showed a convex surface into the pericardium. (2) Podocytes were distributed in the wall of the anterior oblique vein, and the podocytes faced the renopericardial canal (Pirie and George, 1979). It is supposed that the hemolymph fluid was filtered through the lobular wall into the vein, and then Mn²⁺ was concentrated in the kidneys. However, as shown in the sagittal image of the left kidney (Movie 3) and in Fig. 7, the enhanced signal was observed in the long axis of the kidney, but it did not start to increase from the middle of the kidneys. Therefore, the distal part of the renopericardial canal might also function as the kidneys do, and a large surface area might be necessary to reabsorb water and electrolytes, etc. We admit the possibility of back diffusion of Mn²⁺ from the kidney, because we could not detect any valve structure in the renopericardial canal or the convoluted funnel (renopericardial funnel) as reported by Pirie and George (1979).

The kidney is composed of a series of highly branched lobules with a single layered columnar epithelium with a brush border, basal nuclei, infolded basal membranes and many membrane-limited granules (Pirie and George, 1979). The kidney is one of the marked organs that accumulate metals such as cadmium (Cd), copper (Cu), mercury (Hg), zinc (Zn) and Mn (Bayne, 1976). Using isolated lipofuschin granules of Mytilus edulis, stability constants (K_M) of Cd²⁺ and Zn²⁺ were determined to be approximately 5 (George, 1983). Because the K_M of these ions is smaller than that of the strong chelating compounds, such as EDTA (K_M=16.5), the binding of Cd²⁺ and Zn²⁺ might be reversible depending on the concentration of Cd²⁺ and Zn²⁺ (George, 1983). For example, the fraction of binding metal for K_M=5 is estimated as 50% at a metal concentration of 10 µmol l⁻¹ (Seo et al., 2013). In general, the K_M for Mn²⁺ is smaller by 1–2 compared with the K_M for Cd²⁺ or Zn²⁺ (Dojindo Molecular Technologies, Kumamoto, Japan). Therefore, the K_M for Mn²⁺ could be approximately 4–3, and chelators could not bind Mn²⁺ effectively at a [Mn²⁺] of less than 100 µmol l⁻¹. In this study, the M. galloprovincialis kidney did accumulate Mn²⁺ in 3 µmol l⁻¹ seawater, and the maximum concentration of Mn²⁺ was 200 µmol l⁻¹, where the Mn²⁺ concentration in the seawater was higher than 10 µmol l⁻¹ (Fig. 4). When the kidneys accumulated Mn²⁺ at concentrations of 200 µmol l⁻¹, the epithelial cells of the kidneys would uptake Mn²⁺ easily, so the lipofuschin granules could bind Mn²⁺ effectively. In the mussel, the toxic concentration of heavy metals in seawater is approximately 100 µmol l⁻¹ (Bayne, 1976). In the frog or rat, cardiac function decreased at Mn²⁺ concentrations of approximately 100 µmol l⁻¹ (Seo et al., 2011; Seo et al., 2013; Yang et al., 2006). Therefore, 200 µmol l⁻¹ Mn²⁺ might be the maximum concentration at which the kidneys can maintain normal function.

The accumulation of metals in the mussel has been used for monitoring pollution, and the kidney is typically one of the marked organs. (Bayne, 1976; Julshamn et al., 2001). Typical Mn²⁺ contents in the kidney was reported as 6 µg g⁻¹ tissue, which is approximately 100 µmol l⁻¹ (Julshamn and Andersen, 1983). Because we may expect an increase in R₁ by 0.6 s⁻¹, we could detect Mn²⁺ pollution by T_1w-MRI of the mussel and could also estimate the Mn²⁺ concentration from the R₁ of the kidney.

The M. galloprovincialis kidney can concentrate Mn²⁺. In other words, the accumulation of Mn²⁺ increases the toxicity to epithelial cells in the kidneys. As mentioned above, the lipofuschin granules contain the detoxication mechanisms of the mussel. We should also point out that the excretion of the urine from the kidneys is also a useful method to prevent the toxicity of the heavy metal. Anatomically, the kidneys are embedded deeply in the soft tissue, and there is no muscle layer in the wall of the kidney (Fig. 1). Therefore, at first, no one could imagine that the kidneys could excrete the urine instantaneously. As shown in Fig. 6, the right kidney cleared the urine after 58 min, and both kidneys emptied when the seawater in the chamber was replaced. Closure of the shells was associated with both cases where urine excretion was observed (Movie 2-3). These results suggest that M. galloprovincialis could excrete urine voluntarily. The presence of muscular tissue in the excretory pore suggested a neural control system (Fig. 1C). However, there are no muscular tissues around the kidneys. Therefore, although the kidney could not contract itself, internal pressure could be increased, possibly by the interstitial pressure, which may be due to closure of the shell by the adductor muscles, and shrinkage of soft tissue by contraction of the retractor muscles in order to minimize residual urine in the kidney (13%). However, as shown in Fig. 7 and Movie 2-3, the shells did not move in seven cases of excretion of the urine. Even when the residual urine in the kidney increased (35%, P<0.01), closure of the shells might not be essential for urine excretion. In contrast, in some mussels, this voluntary excretion of manganese ion from the kidney occurred repeatedly within 2 h, and the timing of excretion was not synchronized in the kidneys on both sides. It is unclear whether this is a physiological process or is due to the toxic effect of the manganese ion. In this study, we did not detect the details of the excretory process because 3D MR images were detected every 3.4 min. Further studies are necessary to document the voluntary excretion of the urine of M. galloprovincialis.

In summary, we (1) used T_1w-MRI imaging in M. galloprovincialis following Mn²⁺ exposure, and detected accumulation of Mn²⁺ in the urine of kidneys, and (2) detected Mn²⁺ accumulation at a concentration of 3 µmol l⁻¹ Mn²⁺ in the seawater, and the maximum Mn²⁺ concentration level in the urine of kidneys was 200 µmol l⁻¹. (3) Using 3D T_1w-MRI, the changes in Mn²⁺ concentration were measured by duration in minutes, and we found that M. galloprovincialis can voluntarily excrete the urine. Thus, it is expected that there is an additional explanation for the large range in metal concentrations in the kidneys of M. edulis (Lobel et al., 1991). Judging from these results, MRI is a useful technique that holds promise for the future investigation of the function of the kidneys in the mussel.

We offer our sincere thanks to Drs T. Okutani, K. Ohishi and T. Maruyama for providing helpful comments. We would also like to express our thanks to Drs D. Gross, V. Lehman and T. Oerther (Bruker Biospin), as well as Ms Y. Imaizumi-Ohashi and Ms M. Yokoi-Hayakawa (DSUM) for their technical assistance. We must also thank Prof. S. Kojima (AORI, UT) for his helpful suggestions and encouragement to E.S.