INTRODUCTION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

OUR CURRENT UNDERSTANDING of urine concentration through countercurrent multiplication is based almost solely on data for the kidneys of mammals. Some of what has been learned about the renal countercurrent mechanism of the mammalian kidney has been derived from mathematical models and computer simulations (23). Because there are some differences in the morphology of the looped nephrons of the avian kidney and those of mammals, it may not be appropriate to apply existing modeling studies of the mammalian kidney to the avian kidney. These differences, plus the lack of quantitative morphological data for the medullary region of the avian kidney, prompted the current investigation. The aim of the present study was to determine the gradation in loop of Henle length, the coalescence of collecting ducts, and the surface area occupied by the different tubule elements at successive levels from the base to the apex of avian medullary cones.

Aside from mammals, birds are the only group of vertebrates that conserve body water by producing urine osmotically more concentrated than the plasma from which it is derived. However, this ability by birds to concentrate urine is limited compared with that of mammals. Typically, water-deprived birds produce urine that is 2.0-2.5 times more concentrated than plasma, whereas some mammals are capable of producing urine that is 20-25 times more concentrated than plasma (11, 17, 18). To explain these differences, the variation in the organization of the kidneys of birds and mammals should be examined.

Gross kidney structure in mammals ranges from the simple unipapillate of small mammals to the compound multirenculate kidneys found in large marine mammals (10). Despite this range in structure the basic organization of the mammal kidney is maintained and consists of a central medulla that is surrounded by an outer cortex. In contrast, the avian kidney consists of a series of lobules each containing a cortex and a medulla that is shaped like a cone (the medullary cones; Refs. 3, 5, 20). The medullary cones (renal medulla) are not divided into inner and outer regions, as is the case for the kidneys of some mammals.

Although differences in the gross morphology of the renal medulla between mammals and birds are apparent, both medullae contain nephrons with a loop of Henle. Although all nephrons within the mammal kidney contain a loop, in birds only a small percentage of nephrons (10-30%) contain a loop of Henle (i.e., looped nephrons) (4). The remaining large population of nephrons in the avian kidney does not have a loop of Henle. The cortex of the avian kidney is made up of the proximal tubules of the looped nephrons and the large number of nephrons without a loop of Henle (loopless nephrons). In mammal kidneys, the numbers of collecting ducts decrease with distance down the renal medulla. The same is true in the medullary cones of the avian kidney. Moreover, in the avian kidney, the collecting ducts from the loopless nephrons merge in the medullary cones with those of the looped nephrons (3).

Although all the nephrons in mammal kidneys have a loop of Henle, the structure of the loop depends to a degree on the point at which the hairpin loop is formed. The very short nephrons have short, thin descending segments and thick epithelial cells may form the hairpin loop. For the longer nephrons, thin epithelial cells form the hairpin loop (1, 2). In the avian kidney, the hairpin loop of all the looped nephrons consists of thick epithelial cells, i.e., the epithelium of the avian loop of Henle always thickens prior to the hairpin turn (21).

Concentration of the urine by the mammalian kidney is achieved by the parallel arrangement of nephron elements in the renal medulla that function as a countercurrent multiplier. In this system, the tubule fluid is diluted by the active reabsorption of solute (sodium chloride) from the ascending thick limb of Henle without the simultaneous removal of water from the tubule. Deposition of solutes (sodium chloride and urea) in the medullary interstitium creates an environment that is hyperosmotic to the plasma. As the tubule fluid descends in the collecting ducts, antidiuretic hormone facilitates the movement of water to the hyperosmotic interstitium. The function of the thin ascending limb of Henle in this countercurrent multiplier system is unclear (9, 23). The avian kidney has a similar parallel arrangement of the loops of Henle and collecting ducts in the medullary cones, making it likely that it also concentrates urine by a countercurrent multiplier system. However, the intramedullary osmotic gradient of the medullary cones consists of a single solute, sodium chloride (12, 25). The avian countercurrent multiplier system is further complicated and possibly compromised because the collecting ducts in the medullary cones receive fluid from nephrons in the renal cortex that do not have loops of Henle (3). The fluid from these nephrons may be hypotonic or at best isotonic to plasma.

Several mathematical models have been proposed in an attempt to explain the way the mammalian countercurrent multiplier system functions (22, 26-28). In contrast, only one modeling study has examined the countercurrent multiplier system within the avian kidney (16). A second, less quantitative model was proposed for the avian countercurrent multiplier system (19). In this study, the authors used available anatomic and physiological data to propose a model based on a cascade in the length of the loops of Henle with distance down the medullary cones and the observation that sodium chloride is the only solute in the medullary osmotic gradient. However, it did not account for variation in the length of the thin descending segment or the presence of a thick prebend segment to the loop of Henle.

Like mammals, the loop of Henle in birds varies in length among nephrons; however, the pattern of loop formation with depth down the medullary cone has not been examined. Two studies on mammals (one on the rat, the other on the rabbit) have examined the pattern of nephron looping as a function of medullary depth (14, 24). The pattern of loop formation in the medulla may be important to understanding how renal countercurrent multiplier systems operate. Differences in the number of nephrons as a function of medullary depth have been examined in mammals by Knepper et al. (14). Changes in number of nephrons as a function of medulla depth could be responsible for considerable differences in the quantities of solute and water transported and thus urine concentrating ability (14). The study on the avian kidney reported in the present paper is analogous to that of Knepper et al. (14).

The present study was undertaken in anticipation that the results would lead to a better understanding of the urine concentrating mechanism of the avian kidney. Furthermore, the data should be useful in constructing mathematical models and computer simulations of the system.

	MATERIALS AND METHODS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

Gambel's quail were collected under state license using a funnel box trap in the vicinity of Tucson, Arizona. This species was chosen because of the large database that exists on the nephron population of its kidneys (3, 6, 7, 21, 29).

Four birds were euthanized with intraperitoneal injections of pentobarbital sodium (0.5 ml of 50 mg/ml stock). The abdomen of each bird was opened by a single medial ventral incision, and the viscera was removed to expose the kidneys. The dorsal aorta was cannulated just cranial to the kidneys, and blood was flushed from the kidneys with phosphate buffer (pH 7.4). To ensure optimum fixation of the renal medulla, the kidneys were perfused-fixed with half-strength Karnovsky's fixative at an osmolality of 750 mosmol/kgH₂O. Both kidneys were removed from the synsacrum and stored in the perfusion fixative overnight at room temperature.

For each animal, all of the medullary cones were dissected from both kidneys. They were placed in a petri dish, and cones were selected at random for analysis. The cones were embedded in paraffin wax and processed routinely for light microscopy. Serial transverse sections were cut through the length of the cones at a thickness of 5 µm. The sections were mounted on glass slides and stained with hematoxylin and eosin. A total of 11 cones that ranged in length from 950 to 3,350 µm from the eight kidneys of four birds were examined (~3 from each bird). The mean length of all cones was 1,654 ± 738 µm. On average, a single Gambel's quail kidney contains 48 cones, thus we sampled ~3% of the cones in the eight kidneys.

The histological sections were sampled at 50-µm intervals (every tenth section) for the length of the cones. Sampling began at the base of the cones, which corresponded to the corticomedullary boundary and continued to the apex (tip) of the cones (Fig. 1). A Sony charged coupled device camera mounted on an Olympus BH-2 microscope was used to project the sections onto a video monitor. The images were captured using a Targa True Vision frame grabber and analyzed using an image analysis program (Sigma Scan; Jandel, San Rafael, CA). At each level, the numbers of tubule elements were counted (thin and thick limbs of Henle and collecting ducts). For each tubule, the inside and outside diameters were measured (Fig. 1). These data were used to calculate the cross-sectional cell surface area of each tubule using the formula for calculating the surface area of an annulus [A =  (r₁ + r₂) (r₁  r₂), where r₁  r₂]. These data are presented in Table 1.

View larger version (38K): [in this window] [in a new window]   Fig. 1.   A: schematic showing the division of an avian medullary cone into fractional depths. B: a single cross-section of a cone showing the different tubule types. C: parameters measured for each tubule type. Measurements of the inner and outer diameters were made using imaging software (Jandel Sigma Scan Pro), and the data were used to calculate the cross-sectional surface area of each tubule type. The tubules were treated as an annulus and the formula [A =  (r1 + r2) (r1  r2), where r1  r2] was used to calculate the cross-sectional surface area. D: schematic of an avian looped nephron. Note that the epithelium thickens before the hairpin loop is formed. RC, renal corpuscle; PCT, proximal convoluted tubule; PST, proximal straight tubule; Th DLLH, thin descending limb of the loop of Henle; ALLH, ascending limb of the loop of Henle; Tk DLLH, thick descending limb of the loop of Henle.

In addition to the histology, direct quantitative data on the length of the thin limb of Henle and the prebend segment were generated by dissecting 59 looped nephrons from the medullary cones of an additional five Gambel's quail. Birds were prepared as described earlier; however, instead of perfusion fixation, fresh kidneys were dissected from the synsacrum and immersed in a solution of alcoholic ferric chloride (95 ml ethyl alcohol, 5 ml concentrated HCl, 30 g ferric chloride) overnight at 4°C. This was followed by digestion for 2 h in 20% HCl at 37°C, after which the tissue was placed in cold (4°C) acid ferric chloride (200 mg ferric chloride, 0.2 ml acetic acid, and 100 ml distilled water). The kidneys were rehydrated in water for 4-12 h, and individual nephrons were dissected free using finely drawn glass needles. The nephrons were transferred to a drop of 50% glycerol on a microscope slide, and the outline of the nephrons was drawn with the use of a camera Lucida attached to a microscope. The total length of the loop of Henle and the lengths of the thin limb of Henle and prebend segment were measured using the imaging software previously mentioned.

	RESULTS

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

The renal medulla of birds is composed of a number of small units, the medullary cones. As demonstrated in this study, the cones vary widely in diameter at the base, total length, number of loops of Henle, and in the number of collecting ducts. Because of the large variation in these parameters, it would confer little biological value to present the data as means across the cones. However, and as is typical for the desert quail, the internal organization of the cones is very consistent within a given species. This architecture of the cones can be seen in Fig. 2, which contains scanning electron micrographs of methyl methacrylate casts of the loops of Henle and collecting ducts. Figure 2A clearly shows the pattern of coalescence that occurs in the collecting duct system. This coalescence appears not to be random but occurs at specific nodal points, and it appears that this pattern holds for all the collecting ducts in a cone. Inspection of the data in Table 1 on the number of collecting ducts with medullary cone depth also reinforces this pattern of convergence. For example, the number of collecting ducts in cone three decreases in the following pattern: 19, 12, 8, 4, 2. This dichotomous pattern of coalescence has been observed previously for the desert quail and the domestic chicken (3).

View larger version (129K): [in this window] [in a new window]   Fig. 2.   A: scanning electron micrograph of a methyl methacrylate cast of the collecting ducts in a single medullary cone of a quail kidney. Arrows show points of convergence of collecting ducts. Note that the collecting ducts coalesce at distinct levels within the cone structure. B: scanning electron micrograph of a methyl methacrylate cast of the pattern of nephron loops within an avian medullary cone (Gambel's quail). Arrows show the turning points of looped nephrons. Note that the loops of Henle turn at differing depths within the medullary cone.

Numerical data describing the composition of the eleven medullary cones are presented in Table 1. The data are expressed as a function of fractional depth of the cones with the absolute length of the cones taken as 100% and the corticomedullary boundary as zero. As pointed out previously, there is a great deal of variation in the length of the cones within a kidney of species. This is apparent as we present data on cones that ranged in length from 950 to 3,350 µm. There was also variability in the number of nephrons among the cones. Shorter cones did not always have a smaller number of nephrons and vice versa. This is evident when the relationship between number of loops for each cone and cone length is examined (r = 0.27, Fig. 3).

View larger version (13K): [in this window] [in a new window]   Fig. 3.   The absolute number of nephrons at the base of the cone as a function of total cone length. The absolute number of nephrons was calculated by taking the number of descending and ascending limbs of Henle and dividing by two. The equation describing the regression line is y = 32.1 + 16.2x (r2 = 0.137).

In all cones the number of thick limbs of Henle was the greatest at the base and showed a rapid, exponential decrease toward the apex of the medullary cones (Table 1). This point is shown graphically for one cone (cone 11) in Fig. 4A. In all cases, the decrease in the number of thin limbs of Henle toward the apices of the cones was more rapid than the decrease in number for the thick limbs of Henle (Figs. 4A and 7; Table 1). Moreover, there was little correlation between the number of thin limbs and the length of the medullary cones (r = 0.25, P < 0.05).

View larger version (16K): [in this window] [in a new window]   Fig. 4.   A: fraction of thin descending limbs of Henle and thick ascending limbs of Henle as a function of fractional depth toward the apex of a representative medullary cone. B: fraction of collecting ducts and total number of nephrons (loops of Henle) as a function of fractional depth toward the apex of a representative medullary cone. In both A and B, cone 11 is represented. The number of tubule elements toward the apex of the cone is expressed as a fraction of those present at the corticomedullary boundary. The absolute length of the cone was set at 100%.

The number of collecting ducts was greatest at the base of the cones and showed a rapid decrease toward the tip of the medullary cones (Fig. 4B). At the midpoint of the cones ~75 ± 8.0% of the collecting ducts had disappeared. There was little correlation between the length of the cones and the number of collecting ducts at the base (r = 0.18, P < 0.05).

The structure of the descending limb of the loop of Henle was variable with respect to the length of the prebend thick and the thin descending segments of the loop among nephrons of the same length. The variation in the prebend thick segment was such that little correlation existed between the length of this segment and total loop length (Fig. 5A, r = 0.27). The length of the prebend segment from isolated nephrons ranged from 130 to 770 µm. However, the length of the thin limb of Henle was highly correlated with the length of the loop of Henle (Fig. 5B, r = 0.99). Among all isolated nephrons, the length of the thin limb ranged from 390 to 3,970 µm.

View larger version (15K): [in this window] [in a new window]   Fig. 5.   A: length of the prebend thick segment as a function of loop of Henle length. The equation describing the line is y = 223.9 + 0.07x (r2 = 0.27). The length of the prebend thick segment and the total loop length do not appear to be interrelated. B: length of thin limb of Henle as a function of the total length of the loop of Henle (i.e., the length of the thin limb of Henle and the prebend segment). The equation describing the line is y = 223.9 + 0.93x (r2 = 0.99).

To obtain a measure of the ion transporting area of the epithelia at the levels of the cones analyzed the inner and outer diameters of the tubules were measured. These data were used to calculate the cross-sectional cell surface area of the individual tubules (descending and ascending limbs of Henle and collecting ducts). The formula for the surface area of an annulus was used to calculate the surface area of the tubules [A =  (r₁ + r₂) (r₁  r₂), where r₁  r₂]. At each level, the total cross-sectional cell surface area was calculated for each tubule type. The data are presented in absolute numbers and in relative terms with the measurements at the base of the cones taken as 100% (Table 1, Fig. 6).

View larger version (23K): [in this window] [in a new window]   Fig. 6.   Cross-sectional cell surface area vs. fractional depth toward the apices of the cones for thick and thin limbs of Henle and collecting ducts. The base of the cones was set at zero and the apex at 100%. Data are means ± SE for all cones.

The surface area occupied by each tubule element decreased toward the apices of the cones. However, for the thin limbs of Henles loop, the surface area present at 25% of the distance toward the apices the cones exceeded the area at the base of the cones by 17.5% (Table 1, Fig. 6). Beyond this point, surface area of thin and thick limbs of Henle and collecting ducts decreased in a similar manner as the apex of the cones was approached (Fig. 6).

Data on the number of tubules at each level were regressed against cone depth. The regression equations of the summarized data for all cones are presented in Table 2. These data were converted to log base 2. The exponent from each equation indicates the number of convergences of each collecting duct type as a function of fractional cone depth. For example, at the base of cone 11, there are 26 collecting ducts (Table 1). The exponent for collecting ducts for cone 11 indicates that on average there will be 4.11 convergences along the length of the cones (Table 2). Thus starting with 26 collecting ducts, after one convergence there will be 13, then 6.5, 3.25, and finally 1.6. This last value (1.6) is close to the actual value of two, which is the number of collecting ducts remaining near the apex of this cone (see Table 1).

Based on our observations and the data presented, a two-dimensional diagram is presented depicting the lengths of the loops of Henle as they may appear in the medullary cones (Fig. 7). The data indicate that most of the loops of Henle are rather short and that the lengths of the thin limb and prebend segment vary among nephrons. The length of the thin limb of Henle is greater in longer nephrons, and the length of the prebend segment is independent of the length of the loop of Henle.

View larger version (15K): [in this window] [in a new window]   Fig. 7.   Schematic representation showing pattern of the cascade of loop of Henle lengths that are present in the Gambel's quail medullary cone. Note that the majority of nephrons have a short loop of Henle and that the lengths of the thin descending limbs of Henle and prebend segments vary among the nephrons.

	DISCUSSION

TOP ABSTRACT INTRODUCTION MATERIALS AND METHODS RESULTS DISCUSSION REFERENCES

In this study the general features of the renal medulla of the kidney of the desert quail were described and analyzed. The renal medulla of birds is divided into smaller units than occur for most mammalian kidneys. These units, medullary cones, are made up of loops of Henle, collecting ducts, and vasa recta. The individual cones function in much the same manner as the renal medulla of mammalian kidneys. The goal of the present study was to more clearly determine the architecture of the medullary cones with the hope that the information would lead to a better understanding of the function of the countercurrent multiplier system within the avian kidney.

The cones analyzed in this study ranged in length from 950 to 3,350 µm. The data show that cone diameter is independent of cone length, which suggests that the number of loops of Henle entering a cone is also independent of cone length if it is assumed that a larger diameter would accommodate more loops. It follows that short cones may have a large number of loops and the converse is true for long cones.

One of the basic premises of the countercurrent multiplier theory as it applies to the kidney is that the longer the loops of Henle, the greater the multiplier effects. Therefore, the longer cones (with longer loops of Henle) should be able to conserve more water through the production of hyperosmotic urine. This has the potential of being beneficial, but the output from the longer cones is probably counterbalanced by the product of the shorter medullary cones, as the fluid from all cones eventually mixes in the ureteral branches and the ureter. This may be one factor that limits the maximum concentration of the urine produced by avian kidneys.

Studies on the countercurrent multiplier system of mammals suggest that a cascade in the lengths of the loops of Henle is important in the production of concentrated urine (15, 16). The cascade is produced by a decrease in the number of loops of Henle toward the apices of the renal medulla, which gives the medulla (or cone) a tapered shape. The fluid in the longer descending limbs equilibrates osmotically with the interstitium that has been rendered hyperosmotic by the action of the ascending limbs of shorter loops of Henle. Thus the shorter loops preconcentrate the fluid in the descending limbs of the longer loops of Henle as these limbs traverse the interstitium at the level of the shorter loops. The fluid in the descending limbs of the longer loops is further concentrated by equilibration with the interstitium of the deeper medulla. In modeling studies using data for the rat kidney, nephrons that turn too early or too late in the medulla lower the maximal concentrating ability of the cascade system (15, 16). In the rat the distribution of lengths in the loop of Henle population may be near optimal, allowing the system to work at peak efficiency as predicted by the cascade model (15). Similar simulations should be carried out with the data in the present paper to determine whether the loop population occurring in the quail kidneys allows for the most efficient function of this countercurrent multiplier system.

The single solute hypothesis for urine concentration in birds does not include a role for the prebend segment (19). However, more recently Layton and Davies (16) have hypothesized that the prebend segment may act in a manner similar to the diluting segment and aid in amplifying concentrating capacity. In Gambel's quail, the ultrastructure of this segment is similar to that of the ascending limb in that it is densely populated with mitochondria (21), suggesting that it may play a role in sodium chloride absorption. This in turn may aid in developing an interstitial concentration gradient to enabling the passive reabsorption water from the collecting duct network thus producing hyperosmotic urine. Our findings indicate that this segment may be of a substantial length (up to 770 µm); hence the possible importance of this region of the nephron to the function of the countercurrent multiplier system should not be overlooked.

The observation that the number of loops of Henle in the medullary cones of the quail kidney decreases toward the tip of the medulla is consistent with those for the renal medulla of the rat and rabbit kidneys (13, 14). There is no region of the avian medullary cone that corresponds to the inner medulla that is present in most mammal kidneys (5). The data for the rat and rabbit kidneys show a decrease in number of thick ascending limbs of Henle and collecting ducts from the corticomedullary boundary to the boundary of the outer stripe of the outer medulla (13, 14). Thus it would appear that the avian medullary cones are structurally similar (analogous) to the outer medulla of mammal kidneys. The question remains as to whether these two areas function in a similar manner in avian and mammalian kidneys. Simulation based on the data presented here (and other data from the literature) may clarify this point.

In the medullary cones, the number of collecting ducts decreases as a function of length, but the inner and outer diameters of individual ducts increase with distance down the cones. The result is that toward the cone apices, there are fewer, larger collecting ducts. A similar pattern is found with the collecting ducts of rat and rabbit kidneys (14). Although the inner and outer diameters of individual collecting ducts increase with distance down the cones, the decrease in number of collecting ducts is such that the total cross-sectional area decreases with cone depth. The data for cross-sectional cell surface area of the thin and thick limbs of the loop of Henle follow the same pattern as that for the collecting ducts. The decrease in cross-sectional cell surface area seen for the limbs of Henle is in concert with the functional cascade that is apparently necessary for the efficient function of a renal countercurrent multiplier system. The increase in size (luminal diameter) of individual collecting ducts with distance down the cones may be related to the excretion of uric acid by the avian kidney. Uric acid is excreted as spheres that grow larger with transit through the renal tubules. They can reach a size of 15 µm as the late collecting ducts are reached (8).

Based on our data we present a qualitative representation of the morphology for the thin and thick limbs of Henle within the quail medullary cone (Fig. 7). This representation fulfils some of the parameters of our data: 1) the numbers of thin and thick limbs of Henle decrease with distance down the cone, 2) the most rapid decrease in the number of limbs occurs in the early part of the cone (near the base), 3) the rate of decrease of thin limbs of Henle is greater than for the thick limbs, and 4) the length of the prebend (thick) segment varies among adjacent nephrons.

In summary, the avian medullary cone presents a structural organization that is similar to the outer medulla of the mammalian kidney. Studies on the avian kidney that lacks the inner medulla may aid in determining the function of this region in the mammalian kidney.

Perspectives