- Email: [email protected]
Ammonia Transport in the Proximal Tubule In Vivo
Ammonia Transport in the Proximal Tubule In Vivo L. Lee Hamm, MD, and Eric E. Simon, MD • Studies were performed to characterize the determinants of proximal tubule ammonia entry (and retention) In vivo. Rat proximal tubules were studied in vivo using in situ microperfusion. In both normal animals and animals with metabolic acidosis, increasing luminal flow rate significantly enhanced luminal ammonia entry. In contrast, luminal pH was not as important In determining ammonia entry. Analysis of the levels of luminal NH3 in these studies was not consistent with simple diffusion equilibrium of NH 3• In animals with chronic metabolic aCidosis, additional studies demonstrated that inhibition of the Na +·H + exchanger had no direct effect on luminal ammonia entry. However, studies of ammonia efflux from tubules perfused with 10 mmol/L ammonia demonstrated signifi· cant transport of both NH3 and NH4 +• Studies of luminal glutamine deamidatlon via 'Y.glutamyltransferase in con· trol conditions did not indicate a significant role for luminal ammonlagenesis in the superficial proximal tubule in vivo. These and other recent studies of proximal tubule ammonia transport significantly modify the traditional diffusion equilibrium (of NH 3) model of ammonia transport. Luminal flow rate is an important determinant of luminal ammonia entry. Transport of NH4 +, both into and out of the tubule lumen, represents a major component of total ammonia transport. © 1989 by the National Kidney Foundation, Inc. INDEX WORDS: Ammonia; acid·base homeostasis; glutamine; proximal tubule; 'Y·glutamyltransferase; metabolic acidosis; amiloride; microperfusion.
U
RINARY AMMONIA excretion represents the predominant mechanism of renal acid excretion and the proximal tubule is the principal site of ammonia production. (The term "ammonia" will be used to indicate total ammonia, ie, both NH3 and NH4 +.) Recent studies have clearly demonstrated that ammonia transport in the kidney involves the transport of NH4 + in addition to nonionic diffusion of NH3'! Therefore, the entry and retention of luminal ammonia in the proximal tubule may depend on factors other than the passive distribution of NH 3. The purpose of the present study was to examine the determinants of proximal tubule ammonia entry (and retention) in vivo. The determining factors that are examined in the CUfrent report include luminal flow rate, luminal pH, chronic metabolic acidosis (with control ofluminal flow and pH, and after inhibition of Na + - H + exchange), efflux of NH3 and NH4 +, and intraluminal production of ammonia from glutamine. Studies were performed using the rat proximal tubule in vivo since (1) acid-base transport in this species has been best characterized and (2) in vitro experiments may neglect important contributions of such in situ conditions as interstitial ammonia, ammoniagenic substrates, and the hormonal milieu. Hopefully, such in vivo experiments will not only provide information on the regulation of proximal tubule ammonia entry but will also provide information on the relative importance of various mechanisms of ammonia transport in situ which have been described in vitro.
METHODS All of the studies use standard techniques of in vivo microperfusion of superficial proximal tubules of male Munich-Wistar rats. Chronic metabolic acidosis was induced in some animals using NH.Ci administration; this resulted in a mean plasma [HC0 3-] of 13 mmollL in these animals. Bicarbonate concentrations in the perfused and collected fluid were measured by microcalorimetry and total ammonia concentrations in perfusate and collected fluid were measured using an enzymatic fluorometric assay.2 Glutamate in the perfusate and collected fluid was measured in some experiments using a new fluorometric assay that we developed. All of the perfusion solutions were artificial solutions with variations in [HC0 3-] accomplished using reciprocal changes in [CIT In the efflux studies, 10 mmollL NH.Ci was added to the initial perfusion solution; in all other studies, the initial perfusate had no added ammonia. All solutions were gassed with 10% CO 2, The details of the experimental methods have been described previously.3-5 Standard errors and statistical analysis are provided in the original publications but are not repeated for the purpose of simplification; references to significant differences refer to P < 0.05 by analysis of variance or Student's t test.
From the Departments of Medicine, Washington University and The Jewish Hospital of St Louis at Washington University, St Louis, MO. Supported by National Institutes of Health Grants No. AM35023, AM-34394, and 2S07RR5491-25. L. Lee Hamm is an Established Investigator of the American Heart Association. Address reprint requests to L. Lee Hamm, Box 8126, Renal Division, Washington University School of Medicine, 660 S Euclid, St Louis, MO 63110. © 1989 by the National Kidney Foundation, Inc. 0272-638618911404-0003$3.0010
American Journal of Kidney Diseases, Vol XlV, No 4 (October), 1989: pp 253-257
253
254
HAMM AND SIMON
Table 1.
Ammonia Transport
Perfusate Group
Vi (nLlmin)
HC03 (mmoIlL)
15 30 15 30
5 5 25 25
15 30 15 30
Collected Fluid
Fluxes
HC03 (mmoIlL)
AMM (mmollL)
NH3 (,LmoIlL)
HC03 Reabs
AMM Entry
0 0 0 0
5.8 4.1 9.5 15.4
0.81 0.63 0.61 0.56
3.00 1.89 3.54 6.34
5.8 21 125 186
4.2 6.7 3.1 6.5
5 5 25 25
0 0 0 0
2.8 2.7 8.7 13.2
1.53 1.44 1.11 0.9
3.07 2.44 6.11 7.64
29 47 147 220
12.9 20.1 7.9 14.7
30 30
5 25
0 0
3.9 18
1.27 0.71
3.30 8.12
21 148
23.2 10.5
50 50
5 25
10 10
5.5 13.3
4.2 2.8
AMM (mmoIlL)
Normal
1 2 3 4 CMA 5 6 7 8 CMA + AMI 9 10 Efflux studies 11 12
Efflux
-136 -179
NOTE. Vi refers to the perfusion rate; HC0 3 Reabs (reabsorption) and AMM (total ammonia) entry or efflux are given in pmollmm/min. NH3 is the mean collected fluid [NH3J calculated from the collected fluid bicarbonate, total ammonia, and an assumed cortical pC0 2 = 60 mm Hg. CMA refers to animals with chronic metabolic acidosis. AMI refers to tubules perfused with an amiloride analog described in the text.
RESULTS
Most of the results to be discussed are given in Table 1 and represent data from our prior published 3·5 and ongoing projects. Examining the effect ofluminal flow rate, in both normal and acidotic animals, ammonia entry is significantly higher at higher flow rates. With the 5 mmol/L HC0 3- perfusate, higher flow rates (45 nLimin) were required to demonstrate this statistically. 3.5 Luminal pH (reflected by perfusate and collected fluid bicarbonate concentrations) was found to have only a small effect. This is seen by comparing ammonia entry at the same flow rate with the two different perfusate bicarbonates. In fact, in the experiments shown in Table 1, the reductions in ammonia entry with higher luminal pH (and bicarbonate) are only significant when the mean luminal bicarbonate concentration is above plasma bicarbonate concentration (group 8 compared with group 6 in Table 1). Note that in contrast to ammonia entry rates, collected fluid total ammonia concentrations did not vary greatly with either flow rate or pH in either normal or acidotic rats. However, collected NH3 concentrations did differ significantly among the various groups, a finding not predicted by simple diffusion equilibrium of NH 3. Chronic metabolic acidosis obviously increased ammonia entry: with each condition listed, ammo-
nia entry rates were more than twice the rate in normal rats. To test the in vivo importance of the proximal tubule apical Na +-H+ exchanger in metabolic acidosis, experiments (groups 9 and 10 compared with groups 2 and 4, respectively) were performed using a perfusate containing a potent amiloride analog, 5-(N-ethyl-N-isopropyl) amiloride (kindly provided by Dr E.J. Cragoe, Lansdale, PA). Despite significant inhibition of bicarbonate reabsorption, ammonia entry was not significantly altered by the amiloride analog. The small and statistically insignificant fall in ammonia entry with the 25 mmol/L HC0 3- perfusate is attributable to the increased luminal [HC0 3-1. To characterize the efflux of ammonia from the proximal tubule lumen, studies (groups 11 and 12) were performed with 10 mmollL ammonia added to the perfusing solutions. The large efflux of ammonia with both 5 and 25 mmol/L bicarbonate perfusates is consistent with a high permeability to NH 3, NH4 +, or both. Efflux was only modestly higher from the high bicarbonate perfusate, despite a much greater mean NH3 concentration along the lumen (log mean luminal [NH31 = 75.2 ± 4.5 /-tmol/L with the 25 mmol/L bicarbonate perfusate versus 25.1 ± 2.3 /-tmol/L with the 5 mmollL bicarbonate perfusate). This observation
AMMONIA TRANSPORT IN THE PROXIMAL TUBULE
and further analysis4 imply a large component of NH4 + transport. Other experiments (not shown in Table 1) were performed to evaluate the contribution in situ of luminal ammoniagenesis via ,,-glutamyltransferase. Although more of the luminal enzyme is present in the more distal proximal tubule, the S2 segment studied clearly has significant amounts (present data and prior studies). Thbules were perfused with 10 mmollL hippurate (to activate the enzyme), 10 mmollL phenylalanine and aspartate (to inhibit reabsorption of any glutamine and glutamate, respectively). Filtered glutamine is normally reabsorbed proximal to the site of most of the luminal enzyme; therefore, a glutamine (and glutamate)-free perfusate was used. Since diffusion into the lumen of glutamate per se was minimal (demonstrated in other experiments), glutamate measured in the collected fluid represents luminal entry and deamidation of glutamine. With these conditions (which should maximize luminal ammonia production) glutamate accumulation was 1.5 ± 0.3 pmollmm/min, compared with ammonia entry of approximately 10 pmollmmlmin; inhibition of the enzyme eliminated most of the glutamate accumulation. Thus in this segment of the nephron in nonacidotic rats, luminal ammoniagenesis (via ,,-glutamyltransferase deamidation of glutamine) does not appear to contribute importantly to ammonia appearance. DISCUSSION
These studies demonstrate that both luminal flow rate and luminal pH influence proximal tubule ammonia entry. The results demonstrate that increasing flow rate in a physiologically relevant range increases total ammonia entry under all conditions studied; the potential mechanisms of this effect will be discussed below. Luminal pH also altered ammonia entry in both normal rats and acidotic rats. However, the effect of luminal pH was actually small compared with the effect of luminal flow rate, and the influence was only significant in the unphysiologic circumstances in which luminal pH values were greater than plasma pH in both normal (data not shown; see reference 3) and acidotic rats (group 8 compared with group 6, Table 1). As expected, ammonia entry is higher in chronic metabolic acidosis at all levels of luminal pH and flow rate; however, both luminal flow nd pH still influence ammonia entry during metabolic
255
acidosis. Both increased proximal tubule flow rate and decreased luminal pH may occur in metabolic acidosis and stimulate ammonia entry. Both of these factors should increase ammonia entry in acidosis, complementing the increased metabolic production of ammonia. The experiments with NH 4Cl in the perfusate demonstrate another important determinant of proximal tubule luminal ammonia concentrations: efflux of ammonia from the lumen. The high apparent permeability of the proximal tubule to ammonia (and consequent backflux) should serve to limit concentrations of luminal ammonia. Hence, proximal tubule ammonia levels may be determined by both entry into and efflux from the lumen. Previous free-flow micropuncture studies have demonstrated loss of ammonia between the early and late superficial proximal tubule. 6 The efflux experiments performed in the present study also strongly suggest the presence of both NH3 and NH4 + transport. NH3 transport is implicated by the pH dependent component of efflux, presumably secondary to variations in NH3 concentrations with pH. A relatively high NH3 permeability is also consistent with more direct measurements of proximal tubule NH3 permeability7 and estimates derived from measurements of cell membrane NH3 permeabilities. A significant component of NH4 + transport in the efflux experiments is indicated by the pH independent component of ammonia efflux (see reference 4 for details of this analysis). In fact, detailed analysis4 suggests that NH4 + efflux exceeds NH3 efflux at pH values below pH 7.4. The data on ammonia entry are also consistent with proximal tubule transport of both NH3 and NH4 +. Transport of NH3 is implicated by the influence of luminal pH on total ammonia entry with some of the conditions studied. The effect of luminal pH is unlikely to result from significant changes in intracellular pH; luminlJl pH has only a small effect on intracellular pH. 8 However, variations in luminal pH that resulted in significant variations in luminal [NH31 had only a small effect on ammonia entry, implying a lesser role for nonionic NH3 diffusion than has been traditionally assumed. Because ammonia entry rates can not be easily correlated with apparent NH3 concentrations, a component of NH4 + entry is implicated. What are the mechanisms of NH3 and NH4 + transport? NH3 transport is assumed to occur only
256
by diffusion. NH4 + transport occurs by several mechanisms in the kidney. However, two possibilities will be considered for the proximal tubule: passive NH4 + transport (which could occur via several pathways) and NH4 + movement on the Na +-H+ exchanger. Passive transepithelial NH4 + transport (presumably diffusional) has been demonstrated in the rabbit proximal tubule7 ; our studies of ammonia efflux in the rat proximal tubule are also consistent with diffusional NH4 + transport. The route of transepithelial, diffusional NH4 + transport, whether transcellular or paracellular, has not been studied. NH4 + is known to pass through K + channels in other epithelia; this has not been addressed in the kidney. NH4 + movement on the Na +-H+ exchanger has been demonstrated in rabbit proximal tubule brush border membrane vesicles 9 and in mouse proximal tubules in vitro 10 ; the latter studies 10 suggested a predominant role in luminal secretion of ammonia in proximal tubules of normal mice. The present experiments do not exclude some role of Na +-H+ exchange in luminal ammonia entry; however, the experiments with the amiloride analog suggest that Na +-NH4 + exchange is not the predominant or an obligate component of ammonia entry in vivo in the rat with chronic metabolic acidosis. The presence of an intact peritubular interstitium in vivo may explain the discrepancy with the prior in vitro studies 1o ; the interstitium and/or the lateral intercellular spaces in vivo might have significant concentrations of ammonia. However, other experimental differences (eg, species, acidosis, [Na +] of the solutions) could account for the differing results. The issues raised by these discrepant findings point toward unexplored areas of proximal tubule ammonia transport: on the one hand, the possibility of regulation of ammonia entry by regulation of the Na +-H+ exchanger, and on the other hand, the possibility of an important role of the cortical interstitium. With these consideratons, what are the mechanisms of the increase in ammonia entry with increasing flow rate? Two possibilities, not mutually exclusive, are a primary increase in production and a primary increase in transport. First, increased ammonia production with increased flow could account for the increased entry. Such a phenomena has recently been described in mouse proximal tubules in vitro l l ; however, these studies demonstrating increased ammonia production with increased flow only compared very low flows (0 to 8 nL/min) with a single physiologic flow rate (22
HAMM AND SIMON
nLlrnin). If our results in vivo represent increased production with increased luminal flow, then the range of effective flow rates extends up to at least 45 nLlrnin (data not shown in Table 1: with 5 mmollL HC0 3- perfusate, ammonia entry at 45 nLlrnin was 10.1 and 37.8 pmol/mmlmin in normal and acidotic animals, respectively). If increased production causes the increased entry with increased flow, the mechanism of the entry into the lumen still needs to be defined. The second possibility for increased entry with increased flow is an increase in the transport of ammonia per se; an increase in transport of ammonia could be primary or secondary to increased production. The proportionate increase in ammonia entry with increasing flow rate would be a predictable finding with diffusion equilibrium of either NH3 or NH4 + from the cell or interstitium. However, the extreme variations in NH3 and lack of correlation with changes in flow rate would seem to argue against diffusive entry of NH3 accounting for the increased entry with increased flow. Increased NH4 + entry could account for the increased entry with increased flow; increased NH4 + exchange for Na + on the Na +-H+ exchanger could account for the flow related changes of ammonia entry. However, our arniloride analog experiments weigh against the necessity of this mechanism. However, some mechanism of diffusive NH4 + entry would be entirely consistent with the observed pattern of variations of collected fluid [NH4 +]: a small decline in concentrations with increasing flow rate. In sum, the present experiments have demonstrated some of the important determinants of proximal tubule ammonia entry in the rat in vivo; these are (in approximate order of importance): the chronic acid-base status of the animal, luminal flow rate, and luminal pH. Luminal deamidation of secreted glutamine does not appear to contribute greatly to luminal ammonia. These studies also suggest some characteristics of the mechanisms of ammonia transport in the proximal tubule in vivo. NH 3 diffusion appears to be less important than other mechanisms. Ammonia transport in the proximal tubule involves both entry and efflux from the lumen. In chronic metabolic acidosis, Na +-NH4 + exchange does not appear to be an obligate or predominant mechanism of ammonia entry. Other mechanisms of NH4 + entry may exist in the proximal tubule. In vivo, the role of the interstitium in ammonia transport and the role of production in the increase of ammonia entry with increased flow remain undefined.
AMMONIA TRANSPORT IN THE PROXIMAL TUBULE
257
REFERENCES 1. Hamm LL, Simon EE: Roles and mechanisms of urinary buffer excretion. Am J Physiol 253:F595-F605, 1987 2. Good DW, Yurek GG: Picomole quantitation of ammonia by flow-through fluorometry. Analyt Biochem 130: 199-202, 1983 3. Simon E, Hamm LL: Ammonia entry along rat proximal tubule in vivo: Effects of luminal pH and flow rate. Am J Physiol 253:F760-F766, 1987 4. Simon EE, Fry B, Hering-Smith K, et al: Ammonia loss from the proximal tubule in vivo: Effects of luminal pH and flow rate. Am J Physiol 255:F861-F867, 1988 5. Simon EE, Merli C, Herndon J, et al: Determinants of ammonia entry during chronic metabolic acidosis. Am J Physiol 256:Fll04-Fl110, 1989 6. Good DW, DuBose TD Jr: Ammonia transport by early
and late proximal convoluted tubule of the rat. J Clin Invest 79:684-691, 1987 7. Garvin JL, Burg MB, Knepper MA: NH3 and NH/ transport by rabbit renal proximal straight tubules. Am J Physiol 252:F232-F239, 1987 8. Alpern RJ, Chambers M: Cell pH in the rat proximal convoluted tubule. J Clin Invest 78:502-510, 1986 9. Kinsella JL, Aronson PS: Interaction of NH/ and Li+ with the renal microvillus membrane Na +-H+ exchanger. Am J PhysioI241:C220-C226, 1981 10. Nagami GT: Luminal secretion of ammonia in the mouse proximal tubule perfused in vitro. J Clin Invest 81: 159164, 1988 11. Nagami GT, Kurokawa K: Regulation of ammonia production by mouse proximal tubules perfused in vitro. Effect of luminal perfusion. J Clin Invest 75:844-849, 1985
U
RINARY AMMONIA excretion represents the predominant mechanism of renal acid excretion and the proximal tubule is the principal site of ammonia production. (The term "ammonia" will be used to indicate total ammonia, ie, both NH3 and NH4 +.) Recent studies have clearly demonstrated that ammonia transport in the kidney involves the transport of NH4 + in addition to nonionic diffusion of NH3'! Therefore, the entry and retention of luminal ammonia in the proximal tubule may depend on factors other than the passive distribution of NH 3. The purpose of the present study was to examine the determinants of proximal tubule ammonia entry (and retention) in vivo. The determining factors that are examined in the CUfrent report include luminal flow rate, luminal pH, chronic metabolic acidosis (with control ofluminal flow and pH, and after inhibition of Na + - H + exchange), efflux of NH3 and NH4 +, and intraluminal production of ammonia from glutamine. Studies were performed using the rat proximal tubule in vivo since (1) acid-base transport in this species has been best characterized and (2) in vitro experiments may neglect important contributions of such in situ conditions as interstitial ammonia, ammoniagenic substrates, and the hormonal milieu. Hopefully, such in vivo experiments will not only provide information on the regulation of proximal tubule ammonia entry but will also provide information on the relative importance of various mechanisms of ammonia transport in situ which have been described in vitro.
METHODS All of the studies use standard techniques of in vivo microperfusion of superficial proximal tubules of male Munich-Wistar rats. Chronic metabolic acidosis was induced in some animals using NH.Ci administration; this resulted in a mean plasma [HC0 3-] of 13 mmollL in these animals. Bicarbonate concentrations in the perfused and collected fluid were measured by microcalorimetry and total ammonia concentrations in perfusate and collected fluid were measured using an enzymatic fluorometric assay.2 Glutamate in the perfusate and collected fluid was measured in some experiments using a new fluorometric assay that we developed. All of the perfusion solutions were artificial solutions with variations in [HC0 3-] accomplished using reciprocal changes in [CIT In the efflux studies, 10 mmollL NH.Ci was added to the initial perfusion solution; in all other studies, the initial perfusate had no added ammonia. All solutions were gassed with 10% CO 2, The details of the experimental methods have been described previously.3-5 Standard errors and statistical analysis are provided in the original publications but are not repeated for the purpose of simplification; references to significant differences refer to P < 0.05 by analysis of variance or Student's t test.
From the Departments of Medicine, Washington University and The Jewish Hospital of St Louis at Washington University, St Louis, MO. Supported by National Institutes of Health Grants No. AM35023, AM-34394, and 2S07RR5491-25. L. Lee Hamm is an Established Investigator of the American Heart Association. Address reprint requests to L. Lee Hamm, Box 8126, Renal Division, Washington University School of Medicine, 660 S Euclid, St Louis, MO 63110. © 1989 by the National Kidney Foundation, Inc. 0272-638618911404-0003$3.0010
American Journal of Kidney Diseases, Vol XlV, No 4 (October), 1989: pp 253-257
253
254
HAMM AND SIMON
Table 1.
Ammonia Transport
Perfusate Group
Vi (nLlmin)
HC03 (mmoIlL)
15 30 15 30
5 5 25 25
15 30 15 30
Collected Fluid
Fluxes
HC03 (mmoIlL)
AMM (mmollL)
NH3 (,LmoIlL)
HC03 Reabs
AMM Entry
0 0 0 0
5.8 4.1 9.5 15.4
0.81 0.63 0.61 0.56
3.00 1.89 3.54 6.34
5.8 21 125 186
4.2 6.7 3.1 6.5
5 5 25 25
0 0 0 0
2.8 2.7 8.7 13.2
1.53 1.44 1.11 0.9
3.07 2.44 6.11 7.64
29 47 147 220
12.9 20.1 7.9 14.7
30 30
5 25
0 0
3.9 18
1.27 0.71
3.30 8.12
21 148
23.2 10.5
50 50
5 25
10 10
5.5 13.3
4.2 2.8
AMM (mmoIlL)
Normal
1 2 3 4 CMA 5 6 7 8 CMA + AMI 9 10 Efflux studies 11 12
Efflux
-136 -179
NOTE. Vi refers to the perfusion rate; HC0 3 Reabs (reabsorption) and AMM (total ammonia) entry or efflux are given in pmollmm/min. NH3 is the mean collected fluid [NH3J calculated from the collected fluid bicarbonate, total ammonia, and an assumed cortical pC0 2 = 60 mm Hg. CMA refers to animals with chronic metabolic acidosis. AMI refers to tubules perfused with an amiloride analog described in the text.
RESULTS
Most of the results to be discussed are given in Table 1 and represent data from our prior published 3·5 and ongoing projects. Examining the effect ofluminal flow rate, in both normal and acidotic animals, ammonia entry is significantly higher at higher flow rates. With the 5 mmol/L HC0 3- perfusate, higher flow rates (45 nLimin) were required to demonstrate this statistically. 3.5 Luminal pH (reflected by perfusate and collected fluid bicarbonate concentrations) was found to have only a small effect. This is seen by comparing ammonia entry at the same flow rate with the two different perfusate bicarbonates. In fact, in the experiments shown in Table 1, the reductions in ammonia entry with higher luminal pH (and bicarbonate) are only significant when the mean luminal bicarbonate concentration is above plasma bicarbonate concentration (group 8 compared with group 6 in Table 1). Note that in contrast to ammonia entry rates, collected fluid total ammonia concentrations did not vary greatly with either flow rate or pH in either normal or acidotic rats. However, collected NH3 concentrations did differ significantly among the various groups, a finding not predicted by simple diffusion equilibrium of NH 3. Chronic metabolic acidosis obviously increased ammonia entry: with each condition listed, ammo-
nia entry rates were more than twice the rate in normal rats. To test the in vivo importance of the proximal tubule apical Na +-H+ exchanger in metabolic acidosis, experiments (groups 9 and 10 compared with groups 2 and 4, respectively) were performed using a perfusate containing a potent amiloride analog, 5-(N-ethyl-N-isopropyl) amiloride (kindly provided by Dr E.J. Cragoe, Lansdale, PA). Despite significant inhibition of bicarbonate reabsorption, ammonia entry was not significantly altered by the amiloride analog. The small and statistically insignificant fall in ammonia entry with the 25 mmol/L HC0 3- perfusate is attributable to the increased luminal [HC0 3-1. To characterize the efflux of ammonia from the proximal tubule lumen, studies (groups 11 and 12) were performed with 10 mmollL ammonia added to the perfusing solutions. The large efflux of ammonia with both 5 and 25 mmol/L bicarbonate perfusates is consistent with a high permeability to NH 3, NH4 +, or both. Efflux was only modestly higher from the high bicarbonate perfusate, despite a much greater mean NH3 concentration along the lumen (log mean luminal [NH31 = 75.2 ± 4.5 /-tmol/L with the 25 mmol/L bicarbonate perfusate versus 25.1 ± 2.3 /-tmol/L with the 5 mmollL bicarbonate perfusate). This observation
AMMONIA TRANSPORT IN THE PROXIMAL TUBULE
and further analysis4 imply a large component of NH4 + transport. Other experiments (not shown in Table 1) were performed to evaluate the contribution in situ of luminal ammoniagenesis via ,,-glutamyltransferase. Although more of the luminal enzyme is present in the more distal proximal tubule, the S2 segment studied clearly has significant amounts (present data and prior studies). Thbules were perfused with 10 mmollL hippurate (to activate the enzyme), 10 mmollL phenylalanine and aspartate (to inhibit reabsorption of any glutamine and glutamate, respectively). Filtered glutamine is normally reabsorbed proximal to the site of most of the luminal enzyme; therefore, a glutamine (and glutamate)-free perfusate was used. Since diffusion into the lumen of glutamate per se was minimal (demonstrated in other experiments), glutamate measured in the collected fluid represents luminal entry and deamidation of glutamine. With these conditions (which should maximize luminal ammonia production) glutamate accumulation was 1.5 ± 0.3 pmollmm/min, compared with ammonia entry of approximately 10 pmollmmlmin; inhibition of the enzyme eliminated most of the glutamate accumulation. Thus in this segment of the nephron in nonacidotic rats, luminal ammoniagenesis (via ,,-glutamyltransferase deamidation of glutamine) does not appear to contribute importantly to ammonia appearance. DISCUSSION
These studies demonstrate that both luminal flow rate and luminal pH influence proximal tubule ammonia entry. The results demonstrate that increasing flow rate in a physiologically relevant range increases total ammonia entry under all conditions studied; the potential mechanisms of this effect will be discussed below. Luminal pH also altered ammonia entry in both normal rats and acidotic rats. However, the effect of luminal pH was actually small compared with the effect of luminal flow rate, and the influence was only significant in the unphysiologic circumstances in which luminal pH values were greater than plasma pH in both normal (data not shown; see reference 3) and acidotic rats (group 8 compared with group 6, Table 1). As expected, ammonia entry is higher in chronic metabolic acidosis at all levels of luminal pH and flow rate; however, both luminal flow nd pH still influence ammonia entry during metabolic
255
acidosis. Both increased proximal tubule flow rate and decreased luminal pH may occur in metabolic acidosis and stimulate ammonia entry. Both of these factors should increase ammonia entry in acidosis, complementing the increased metabolic production of ammonia. The experiments with NH 4Cl in the perfusate demonstrate another important determinant of proximal tubule luminal ammonia concentrations: efflux of ammonia from the lumen. The high apparent permeability of the proximal tubule to ammonia (and consequent backflux) should serve to limit concentrations of luminal ammonia. Hence, proximal tubule ammonia levels may be determined by both entry into and efflux from the lumen. Previous free-flow micropuncture studies have demonstrated loss of ammonia between the early and late superficial proximal tubule. 6 The efflux experiments performed in the present study also strongly suggest the presence of both NH3 and NH4 + transport. NH3 transport is implicated by the pH dependent component of efflux, presumably secondary to variations in NH3 concentrations with pH. A relatively high NH3 permeability is also consistent with more direct measurements of proximal tubule NH3 permeability7 and estimates derived from measurements of cell membrane NH3 permeabilities. A significant component of NH4 + transport in the efflux experiments is indicated by the pH independent component of ammonia efflux (see reference 4 for details of this analysis). In fact, detailed analysis4 suggests that NH4 + efflux exceeds NH3 efflux at pH values below pH 7.4. The data on ammonia entry are also consistent with proximal tubule transport of both NH3 and NH4 +. Transport of NH3 is implicated by the influence of luminal pH on total ammonia entry with some of the conditions studied. The effect of luminal pH is unlikely to result from significant changes in intracellular pH; luminlJl pH has only a small effect on intracellular pH. 8 However, variations in luminal pH that resulted in significant variations in luminal [NH31 had only a small effect on ammonia entry, implying a lesser role for nonionic NH3 diffusion than has been traditionally assumed. Because ammonia entry rates can not be easily correlated with apparent NH3 concentrations, a component of NH4 + entry is implicated. What are the mechanisms of NH3 and NH4 + transport? NH3 transport is assumed to occur only
256
by diffusion. NH4 + transport occurs by several mechanisms in the kidney. However, two possibilities will be considered for the proximal tubule: passive NH4 + transport (which could occur via several pathways) and NH4 + movement on the Na +-H+ exchanger. Passive transepithelial NH4 + transport (presumably diffusional) has been demonstrated in the rabbit proximal tubule7 ; our studies of ammonia efflux in the rat proximal tubule are also consistent with diffusional NH4 + transport. The route of transepithelial, diffusional NH4 + transport, whether transcellular or paracellular, has not been studied. NH4 + is known to pass through K + channels in other epithelia; this has not been addressed in the kidney. NH4 + movement on the Na +-H+ exchanger has been demonstrated in rabbit proximal tubule brush border membrane vesicles 9 and in mouse proximal tubules in vitro 10 ; the latter studies 10 suggested a predominant role in luminal secretion of ammonia in proximal tubules of normal mice. The present experiments do not exclude some role of Na +-H+ exchange in luminal ammonia entry; however, the experiments with the amiloride analog suggest that Na +-NH4 + exchange is not the predominant or an obligate component of ammonia entry in vivo in the rat with chronic metabolic acidosis. The presence of an intact peritubular interstitium in vivo may explain the discrepancy with the prior in vitro studies 1o ; the interstitium and/or the lateral intercellular spaces in vivo might have significant concentrations of ammonia. However, other experimental differences (eg, species, acidosis, [Na +] of the solutions) could account for the differing results. The issues raised by these discrepant findings point toward unexplored areas of proximal tubule ammonia transport: on the one hand, the possibility of regulation of ammonia entry by regulation of the Na +-H+ exchanger, and on the other hand, the possibility of an important role of the cortical interstitium. With these consideratons, what are the mechanisms of the increase in ammonia entry with increasing flow rate? Two possibilities, not mutually exclusive, are a primary increase in production and a primary increase in transport. First, increased ammonia production with increased flow could account for the increased entry. Such a phenomena has recently been described in mouse proximal tubules in vitro l l ; however, these studies demonstrating increased ammonia production with increased flow only compared very low flows (0 to 8 nL/min) with a single physiologic flow rate (22
HAMM AND SIMON
nLlrnin). If our results in vivo represent increased production with increased luminal flow, then the range of effective flow rates extends up to at least 45 nLlrnin (data not shown in Table 1: with 5 mmollL HC0 3- perfusate, ammonia entry at 45 nLlrnin was 10.1 and 37.8 pmol/mmlmin in normal and acidotic animals, respectively). If increased production causes the increased entry with increased flow, the mechanism of the entry into the lumen still needs to be defined. The second possibility for increased entry with increased flow is an increase in the transport of ammonia per se; an increase in transport of ammonia could be primary or secondary to increased production. The proportionate increase in ammonia entry with increasing flow rate would be a predictable finding with diffusion equilibrium of either NH3 or NH4 + from the cell or interstitium. However, the extreme variations in NH3 and lack of correlation with changes in flow rate would seem to argue against diffusive entry of NH3 accounting for the increased entry with increased flow. Increased NH4 + entry could account for the increased entry with increased flow; increased NH4 + exchange for Na + on the Na +-H+ exchanger could account for the flow related changes of ammonia entry. However, our arniloride analog experiments weigh against the necessity of this mechanism. However, some mechanism of diffusive NH4 + entry would be entirely consistent with the observed pattern of variations of collected fluid [NH4 +]: a small decline in concentrations with increasing flow rate. In sum, the present experiments have demonstrated some of the important determinants of proximal tubule ammonia entry in the rat in vivo; these are (in approximate order of importance): the chronic acid-base status of the animal, luminal flow rate, and luminal pH. Luminal deamidation of secreted glutamine does not appear to contribute greatly to luminal ammonia. These studies also suggest some characteristics of the mechanisms of ammonia transport in the proximal tubule in vivo. NH 3 diffusion appears to be less important than other mechanisms. Ammonia transport in the proximal tubule involves both entry and efflux from the lumen. In chronic metabolic acidosis, Na +-NH4 + exchange does not appear to be an obligate or predominant mechanism of ammonia entry. Other mechanisms of NH4 + entry may exist in the proximal tubule. In vivo, the role of the interstitium in ammonia transport and the role of production in the increase of ammonia entry with increased flow remain undefined.
AMMONIA TRANSPORT IN THE PROXIMAL TUBULE
257
REFERENCES 1. Hamm LL, Simon EE: Roles and mechanisms of urinary buffer excretion. Am J Physiol 253:F595-F605, 1987 2. Good DW, Yurek GG: Picomole quantitation of ammonia by flow-through fluorometry. Analyt Biochem 130: 199-202, 1983 3. Simon E, Hamm LL: Ammonia entry along rat proximal tubule in vivo: Effects of luminal pH and flow rate. Am J Physiol 253:F760-F766, 1987 4. Simon EE, Fry B, Hering-Smith K, et al: Ammonia loss from the proximal tubule in vivo: Effects of luminal pH and flow rate. Am J Physiol 255:F861-F867, 1988 5. Simon EE, Merli C, Herndon J, et al: Determinants of ammonia entry during chronic metabolic acidosis. Am J Physiol 256:Fll04-Fl110, 1989 6. Good DW, DuBose TD Jr: Ammonia transport by early
and late proximal convoluted tubule of the rat. J Clin Invest 79:684-691, 1987 7. Garvin JL, Burg MB, Knepper MA: NH3 and NH/ transport by rabbit renal proximal straight tubules. Am J Physiol 252:F232-F239, 1987 8. Alpern RJ, Chambers M: Cell pH in the rat proximal convoluted tubule. J Clin Invest 78:502-510, 1986 9. Kinsella JL, Aronson PS: Interaction of NH/ and Li+ with the renal microvillus membrane Na +-H+ exchanger. Am J PhysioI241:C220-C226, 1981 10. Nagami GT: Luminal secretion of ammonia in the mouse proximal tubule perfused in vitro. J Clin Invest 81: 159164, 1988 11. Nagami GT, Kurokawa K: Regulation of ammonia production by mouse proximal tubules perfused in vitro. Effect of luminal perfusion. J Clin Invest 75:844-849, 1985