Kidney anatomy, histology and histometric traits associated to renosomatic index in Gymnotus inaequilabiatus (Gymnotiformes: Gymnotidae)

The Gymnotus inaequilabiatus is a Neotropical fish widely distributed in marginal areas of bays. The aim of this study was to describe the main histological and histopathological traits in the head and exocrine kidney . Here, histometric and structural density techniques were associated with renosomatic index (RSI). The kidney was processed for light microscopy. Lipofuscin and hemosiderin content were visually estimated in the melonomacrophages centers (MMCs). All the biometric body variables were correlated with RSI, especially the kidney weight and gross lesions count. The general architecture of head and exocrine kidney was similar to that described for other teleost species. MMCs were prevalent in both portions and correlated with RSI in the head and exocrine kidney. Granulomatous structures were often observed in both portions; however, they were associated only in the exocrine kidney with RSI. Of all the structures hystometrically estimated, only proximal tubular diameter and thickness, and distal tubular thickness were correlated to renosomatic index. The RSI is an useful biometric variable that represent some physiological and morphological characteristics of kidney in G. inaequilabiatus . These findings may be used in future studies to evaluate the effects of environmental stressors on the renal adaptative physiological process.


Introduction
The kidney is a representative and integrated organ with many vital functions of the organism. As in other vertebrates, the kidney of fish is the primary organ involved in the osmoregulation process and exerts a critical immunological function (Cuesta et al., 2007;Harper, Wolf, 2009). Anatomically, this organ divides into two portions denominated as cephalic or head, and exocrine or trunk kidney. In some species, no division between portions is observed, although morphology and functions are distinct (Takashima, Hibiya, 1995). The head portion consists of lymphohematopoietic tissue connected by an extensive network of vessels and capillaries drained by the renal portal system (Roberts, 2012). The erythroid, lymphoid, and myeloid cells are situated within a stroma 2 e190107 [2] of reticuloendothelial tissue in association with mature granulocytes and agranulocytes cells (Zapata et al., 1996;Roberts, 2012). The exocrine kidney is comprised of nephrons diffusely distributed, which are still intermediated by hematopoietic and lymphopoietic tissue (Larsen, Perkins Jr., 2005;Menke et al., 2011). Although the histological traits of kidney structures are well known in fish the wide variations in their anatomy, histology, and cytology, related to the habitat conditions make this organ play a primordial role in the adaptive and evolutionary processes in fish (Dantzler, 2016). In this regard, few studies have been conducted given the considerable biodiversity of fish species in the Neotropical region, especially from their natural environments.
Organosomatic indices are used to measure organ mass relative to body mass. In association with other body metrics, organosomatic indices have been widely used in ecological and experimental studies (Schlenk et al., 2008). However, the renosomatic index has been poorly reported in association with useful histological biomarkers. This index tracks metabolic variations in response to parasite infections, heavy metal contamination, involvement in physiological adjustment to environmental conditions or seasonality (Resende et al., 2010;Jordanova et al., 2017;Strepparava et al., 2018). On the other hand, the quantification of morphometric measurement allows detection of minimal differences in cells and tissues, favoring the qualitative analysis in addition to responding to the functional and pathological changes in an organ (Oberholzer et al., 1996;Nyengaard, 1999;Hamilton et al., 2014). Therefore, the combined interpretation of this index would permit the assessment of distinct physiological responses from the internal and external individual factors (Dethloff, Schmitt, 2000).
The Gymnotiformes order comprises more than 250 species distributed in 45 Gymnotus spp. genera and represent nearly four percent of all ichthyofauna of the Neotropical region (Fricke et al., 2019). In the Brazilian Pantanal, Gymnotus inaequilabiatus is distributed in marginal areas of bays formed by flooded areas of low and receding waters of rivers that drain into the Paraguay and Parana River, with abundant aquatic vegetation (Resende et al., 2006). The life history and natural habitat of Gymnotus spp. suggest that they adapted to low oxygen rates or have a wide capacity for survival in extreme hypoxic environments, probably by an outstanding physiological adaptive process (Crampton, 1998;Moraes et al., 2002). On this point, renal histomorphometry assessment can play a key role owing to the functional plasticity of this organ (Sakai, 1985;Resende et al., 2010;Hassan et al., 2017;Pelster et al., 2018).
The Gymnotidae family has been studied from the genetic, taxonomic, endoparasitic fauna and ecological standpoints (Willink et al., 2000;Albert et al., 2005;Scacchetti et al., 2011;Galindo et al., 2017). Although these studies are numerous, there are no reports about anatomy, histology, and histopathological features of the kidney and were not even reported biometrics and histometric traits in this species. Therefore, profiling based on a comparative database following the evolution of histometric traits in specimens under natural habitats might suggest G. inaequilabiatus as a potential environmental bioindicator.
This study aimed to characterize and describe the main histological and histopathological finds in the head and exocrine kidney in the G. inaequilabiatus. Additionally, histometric and structural density techniques were applied to provide quantitative subsets associating them with the renosomatic index.

Material and Methods
Samples and biometric data. Adult G. inaequilabiatus specimens (female, n = 39; male, n = 45) from Porto Morrinho,Corumbá,Brazil,(19°46'49"S 57°41'52"W) were acquired from a fishing community. The specimens were transported in oxygenated polyethylene bags to the laboratory and maintained in aquaria supplied with artificial aeration at constant temperature (25.0°C) until euthanasia (2-Fenoxietanol, 2 mL/L), approximately 3-5 days after transport. Then, the fish were weighed (total weight in grams) and measured (total length in centimeters) to estimate the relative condition factor (Kn) according to the formula Kn = W/aL b , where W is the weight of the individual, L is the standard length of the individual, and a and b are the constants from the weight-length relation (Le Cren, 1951). After that, the fish were necropsied for kidney analysis. This organ was carefully removed and weighed to calculate the renosomatic index (RSI = kidney weight (g)/body weight (g) x 100). The gross lesions evaluation was conducted on the dorsal and ventral surfaces, considering the number, category, and distribution (Bruno et al., 2013).
Histological and histochemical analysis. Fragments of the head and exocrine kidney from 20 specimens randomized (n = 16 females and 14 males) were prepared. After fixation in buffered formalin 10% for 24 hours, the samples were dehydrated in ethanol, embedded in Paraplast Plus (Sigma-Aldrich®) and 5 µm sections were then stained with hematoxylin and eosin (H&E). Masson Trichrome (MT) and Periodic Acid Schiff (PAS) staining were prepared for collagen fibers and glycogen cell deposit analyses (Carson, Hladik, 2009). To detect the lipofuscin content in MMCs, fragments of the same samples were fixed in methacarn solution (60% absolute methanol, 30% chloroform, and 10% glacial acetic). Kidney histological sections were incubated for 1 minute in Schmorl solution, containing 75 mL of 1% ferric chloride, 10 mL of potassium ferricyanide, and 15 mL of distilled water. Next, the sections were immersed in an aqueous 1% neutral red solution followed by 1% eosin. To detect the hemosiderin content, sections were incubated in ferrocyanide acid solution, obtained by dissolving 2 g of potassium ferrocyanide in 100 mL of hydrochloric acid solution 0.75 mmol/L, for 15 minutes, and immersed 3 e190107 [3] in aqueous 1% neutral red followed by aqueous 1% eosin staining (Franco-Belussi et al., 2012). Histological changes, where present, were reported according to morphological characteristics.

Histometry and structural volumetric density analysis.
Histometric measures were processed in sections stained by MT. Ten images/section at 400 × magnification, considering at least five structures per image, were used to estimate the renal corpuscle diameter (µm) and area (µm 2 ), the glomerular area (µm 2 ), the diameter (µm), and thickness of the proximal and distal tubular epithelium. Additionally, the following parameters were calculated: Bowman´s space area (µm 2 ; renal corpuscle area -glomerular area), the glomerular area/renal corpuscle area ratio (Rgrc = glomerular area/ renal corpuscle area x 100) and Bowman´s space area/renal corpuscle area ratio (RBrc = Bowman's space area/renal corpuscle area) x 100. The Motic Image Plus 2.0 (Motic Asia) software was used for this analysis.
For the structural volumetric density analysis, images were captured with a Leica DF495 camera coupled to a bright light microscope (Leica DM5500 B). The volume densities were prepared using sections stained by H&E. Digital graticules overlapped ten images at 400 x magnification with 266 intersections. Subsequently, lymphohematopoietic tissue, melonomacrophages centers (MMCs), blood vessels, sinusoidal spaces, and general tissue alterations were estimated in the head kidney. All these structures, in addition to the glomerulus, proximal, and distal tubules, were assessed in the exocrine kidney. Then, the following formula was applied: Dv (%) = (Ps x 100) / Pr, where Ps is the number of intersections counted for each structure, and Pr is the total number of intersections (Freere, Weibel, 1967;Reid, 1980;Rocha et al. 1997). The ImageJ 1.45M software was used for this analysis.
Statistical analysis. Data were tested for normality and homogeneity by Kolmogorov-Smirnov test and the Shapiro-Wilk's W test. Blood vessels and granulomatous lesions from the structural density analysis were arcsine transformed. No differences were observed between females and males (P > 0.05, one-way ANOVA, t-test) in the entire population (n = 89) and that randomized to the histological analysis (n = 20). Then, the overall mean, standard deviation, and percentis (25th and 75th) were predicted. Pearson's correlation estimated linear correlations between biometrical, structural volumetric densities, and histometric measures with RSI. SPSS 17.0 (IBM®) software was used in the analysis.

Results
Biometric variables such as body weight, length, kidney weight, relative condition factor (Kn), and RSI for the whole population, are reported in Tab. 1. The RSI was correlated to all biometric variables. The highest correlations were found with kidney weight and gross lesions count. The kidney of G. inaequilabiatus is situated in the dorsoventral and retroperitoneal region of the coelomatic cavity. It is composed of two cranial branches or lobes, which comprise the head portion of the kidney. These branches are an extension of the exocrine part of the organ. The color was typically dark brown; however, some samples presented a dark red color. The majority of the lesions were rounded with focal distribution, and whitish displaying regular nodules often raised from the surface. Submacroscopic images revealed the MMCs diffusely distributed in both head and exocrine kidney (Fig. 1).
The histological architecture of head and exocrine kidney is similar to other freshwater teleosts (Fig. 2). The parenchymal arrangement is diffusely distributed without anatomical divisions between glomeruli and excretory structures. Lymphohematopoietic cell niches compose the head organ as a continuity of interstitial tissue observed throughout the exocrine segment. In low magnitude, it was possible to see the transition region between exocrine and head kidney. The sinusoidal networks were widely distributed from the capillaries, vessels, and arterioles . In addition to undifferentiated hematopoietic cells, macrophages, lymphocytes, granulocytic cells, and rodlet cells were found in both head and exocrine kidney. The collecting tubules (rarely seen) displayed tall and narrow columnar cells with nuclei basally situated. The MMCs, which present several shapes and sizes, were observed in both head and exocrine hematopoietic tissue. As to other freshwater fish, the proximal tubules present columnar cells with brush border; however, distal tubules do not have this characteristic. The cells of both segments showed similar size, although the proximal tubules were slightly more eosinophilic. Nuclei were basally located in proximal and distal tubular segments, and a lesser amount of central and apical location was also observed . Hyaline degeneration of the tubular epithelium was randomly found, without, however, joining with other degenerative forms. Stannius corpuscle is localized in the middle region of both branches from the head kidney, covered by a capsule of connective tissue that extends into the corpuscular space. The glandular lobules are constituted by numerous granule cells enclosed by sinusoidal capillary plexus (Figs. 2e-f).   Renal corpuscle (glomerulus and Bowman´s capsule) is filled by a parietal epithelium closed to the interstitial cells, proximal and distal tubules. Visceral layer (podocytes), mesangial cells, capillary, and intracapsular space were identified in the glomerulus. Collagenous structures permeate the intracapsular space and outline of all structures (Figs. 3ab). The lipofuscin and hemosiderin histochemical reactions in the MMCs are displayed in Figs. 3c-f. Hemosiderin and lipofuscin were observed more in the exocrine kidney in contrast to the head. Granulomatous structures were often observed in both kidneys. Usually, the MMCs were found surrounding or inside the granulomas. Internally, the granulomas were typically constituted of cell debris, necrotic material, and collagen fibers. Rodlet cells were observed peripherally to MMC aggregates associated with lymphocytes, macrophages, and eosin granulocytic cells (Fig. 4).  The lymphohematopoietic tissue was the most critical structure of the head and exocrine kidney (Tab. 2). In the head portion, it was followed by the sinusoid lumen and MMCs. Blood vessels and granulomatous lesions were poorly observed; however, both showed a higher variation (standard deviation superior to mean). Low linear correlation coefficients for lymphohematopoietic tissue and MMC volumetric density were estimated. In the exocrine kidney, the percentage of proximal tubules was close to hematopoietic tissue. Distal tubules and MMCs were equivalents, and granulomatous lesions were higher in the head kidney. The glomerulus was the least dense structure, representing approximately 0.41-0.63% of all exocrine kidney parenchyma. Granulomatous lesions were more prevalent in the exocrine kidney. Except for renal corpuscle, all the structures correlated to RSI.
The histometric measures dataset is reported in Tab. 3. The glomerular area represented approximately 83.9 % of the renal corpuscle while the Bowman´s space approximately 16.0 %, although a high range has been observed. The proximal tubular diameter and thickness of the epithelium were larger concerning distal tubular measures. Of the all structures measured, only proximal tubular diameter and thickness, and distal tubular thickness were correlated to RSI.

Discussion
The selective habitat and wide-ranging distribution in the Neotropical region make the Gymnotus inaequilabiatus a notable species to explore physiological and adaptive responses concerning its environment and life history. Typically, this species remains in lotic environments submerged in the mud closest to macrophyte vegetation. These conditions are characterized by low oxygen level, which probably induces anatomical and functional adaptations, such as breathing air through accessory airbreathing organs or via gill system (Crampton, 1998;Moraes et al., 2002). Furthermore, fish are subjected to adverse situations such as large temperature fluctuations, food scarcity, and a broad spectrum of pathological agents (Bonga, 1997;Galindo et al., 2017).
Kidney anatomy of G. inaequilabiatus is arranged according to many other Teleostei fish (Takashima, Hibiya, 1995;Holz, Raidal, 2006). The head and exocrine portions are interconnected, with double branches that remain cranioventrally attached at the vertebral column. Dark brown was the predominant color due to their abundant vascular system and MMCs-rich parenchyma. Also, even though a wide variation is commonly reported, the RSI was similar to other fish species (Stoskopf, 2010). White cystic lesions focally distributed and with a firm consistency characterized the gross lesions. These lesions are common in free range fish populations and are associated with a wide variety of causes (Ferguson, 2006;Pavanelli et al., 2008;Bruno et al., 2013).

Tab. 2.
Descriptive statistic values for structural volumetric densities (%) and Pearson. # considering 10 fields per specimen; SD, standard deviation; MMCs, melonomacrophages centers; a include arterioles, venule, and capillaries; b glomerulus with Bowman's capsule; * p < 0.05; ** p < 0.01. To date, there are no anatomical, histological or histopathological descriptions of the kidney in G. inaequilabiatus, which prevent a comparative analysis within the Gymnotiformes Genus. However, the values found of the relative volumetric density from the proximal tubules are similar to the sum of the proximal tubules I and II values reported in the Salmo trutta (Resende et al., 2010), in accordance to the low percentage of renal corpuscles. Both tubular segments were not estimated separately once the distinct morphological features are more clearly recognized by transmission microscopy (Takashima, Hibiya, 1995;Elger et al., 2000). Further restrictive comparisons, including other renal structures, are still speculative due to the limited numbers of stereological studies, although the overall tissue composition is analogous between kidneys of different fish species. Given the wide variety on the morphological and anatomical kidney trends, the RSI was a useful measure since it correlated with several macroscopic and microscopic attributes. This index reflects the proportional sizes of organs relative to the body mass. Many environmental factors influence their variation and indicate the general status of the health of an individual organism (Strepparava et al., 2018). Although we have found a positive linear correlation with relative condition factor, the correlation with the number of gross lesions was high (r = 0.48, Tab. 1). Furthermore, the linear correlation with the volumetric density of MMCs was negative in both head (r = -0.22) and exocrine portions (r = -0.31, Tab. 2). The coefficient measured between gross lesions and RSI suggests that adult or aged specimens with higher indices would tend to sustain more injuries through the organ. The negative correlations with MMCs may be due to these structures remaining aggregated around the granulomatous lesions and moderately dispersed in the renal parenchyma, a feature previously reported to hepatic nematode-cysts granulomas (Galindo et al., 2017). In the present study, MMCs account for a substantial fraction of the volumetric density in both the head and exocrine kidney mostly composed of melanin (Agius, Roberts, 2003;Kumar et al., 2016). Hemosiderin and lipofuscin were promptly identified, although they have not been quantified. In addition to the immune function, MMCs are involved in the endogenous phagocytosis material of exhausted cells, particularly erythrocytes. Thus, these structures in both kidneys may be implicated in a stable iron-recycling function and antimicrobial response (Steinel, Bolnick, 2017;Grayfer et al., 2018). The volumetric density of MMCs is consistent with those reported in Salvelinus fontinalis and Oncorhynchus mykiss, even though the measurement techniques of their content have been different (Schwindt et al., 2006). In contrast with this study, granulomatous lesions do not appear to be an occasional finding once were frequently observed in all samples in G. inaequilabiatus.
The volumetric density of lymphohematopoietic tissue was measured independently between the head and exocrine portions, once both portions were morphologically delimited and microscopically distinguishable. It constituted the highest relative volumetric density percentage of structures evaluated in both portions. However, in the exocrine kidney except for the renal corpuscle, all structures observed were associated with RSI, suggesting that renal volume dynamics depend on subtle pathophysiological adjustments. The nephrogenesis in adult fish seems to be a continuous regenerative process related to fish body size and nephron number (Reimschuessel, 2001;Davidson, 2011). The proximal and distal tubule segments showed highest correlations, and both are functionally active for osmoregulation, urea, creatinine, amino acids and glucose homeostasis (Larsen, Perkins Jr., 2005;Dantzler, 2016). Also, proximal tubules presented typical and irregular eosinophilic hyaline granules, probably in accordance with reabsorption of excess amounts of proteinaceous substances depurated from the glomerulus. This process has been observed under both pathological and healthy fish (Takashima, Hibiya, 1995;Wolf et al., 2015).
The head and exocrine kidney presented an appreciable vascular network. The expressive volumetric density of sinusoidal lumen suggests a similar function to that performed by the spleen, where foreign bodies, such as bacterial, fungal or parasitic fragments are trapped within a fine reticular meshwork (Fergusson, 2006;Roberts, Ellis, 2012). While in the head portion the sinusoidal lumen was prominently configuring meaningful flow irrigation among the lymphohematopoietic cell niches, the blood vessels were prominent in the exocrine portion. However, only blood vessels from the exocrine portion correlated to RSI. Indeed, in most teleost fish, the glomerular blood supply is arterial while the renal tubules have a second supply of blood; the first emerge from the efferent glomerular arterioles, and the second from the venous renal-portal system Holz et al., 2006). Thus, considering the higher density of proximal and distal tubules distributed throughout the tissue, the venous vasculature may be regarded as a notable feature in this species. The interactive, functional, and volumetric density aspects of tubular and parenchymal vasculature retained an expressive account on the RSI estimative.
Histometric measures contribute to the description of the tissue shape and functionality, which varies according to the species characteristic and might express different levels of healthiness (Reid, 1980;Nyengaard, 1999). Proximal and distal tubular measures are related to plasticity and remodeling of epithelial cells in response to reabsorption and excretory function in teleosts fish adapted to freshwater (Oğuz, 2015;Dantzler, 2016). The proximal tubular diameter and thickness, as well as distal tubular thickness, were negatively correlated with RSI. These measures implicate in size-attributes and may modulate this index since both segments represented the highest percentage of renal structures. Our results suggest that lower and narrower tubules would be concentrated in adults and heavier specimens with more significant renosomatic indices.
We described for the first time the morphological baseline in light microscopy of G. inaequilabiatus. The RSI is associated with various functional structures, and their variation accounts significantly for the volumetric density of lymphohematopoietic tissue, as well as of the proximal and distal tubular segments. The prevalence of granulomatous lesions in both kidney portions indicate to us that these act as target-organs for arresting foreign bodies, composing a critical immune response in this species. This particularity, along with MMCs relative density and pigmentary characteristics could address us news essays in order to elucidate the renosomatic index relation to environmental factors.