Kidney International, Vol. 42 (1992), PP. 690—699

Abnormal cytoplasmic pH regulation during activation in uremic neutrophils ANDREW P. HAYNES, IAN DANIELS, CHRISTINE PORTER, JOHN FLETCHER, and ANTHONY G. MORGAN Department of Haematology, Medical Research Centre, and Department of Renal Medicine, City Hospital, Nottingham, England, United Kingdom

Abnormal cytoplasmic pH regulation during activation in uremic neutrophils. Neutrophils isolated from ESRF patients demonstrated abnormal cytoplasmic pH changes after FMLP stimulation; the initial cytoplasmic acidification was absent (P < 0.001 compared to controls)

oxidants generated by the respiratory burst and products released into the phagosome by degranulation [reviewed in 10,

11]. Well-described biochemical changes accompany PMN activation, and these are important for the control of degranulaand the degree of alkalinization enhanced (P < 0.05 compared to controls). This effect was not due to the absence of any of the factors tion and the respiratory burst. These include changes in intraassociated with acidification in normal PMN since superoxide produc- cellular calcium levels and cytoplasmic pH. In this paper we tion was enhanced (P < 0.05 compared to controls) and intracellular report our observations upon the changes in these parameters calcium release was normal. Our observations are not explicable by alterations in the function of the Na:H antiport since the kinetics of accompanying activation in uremic PMN. antiport activation by cytoplasmic pH were not different in uremic and control cells. Other factors must therefore be important in generating the abnormal pH response to chemotactic factors in uremic PMN. Cells from CAPD patients had some degree of initial acidification (P < 0.001 compared to controls and P < 0.05 compared to ESRF) and enhanced alkalinization (P < 0.05 compared to controls). Preincubation of normal PMN in four-hour dwell PDE reproduced the responses of uremic PMN with absent acidification, enhanced alkalinization and enhanced super-

Methods

Subjects

Fourteen patients with end-stage renal failure (ESRF), hence plasma creatinine > 500 LM/liter (mean 719 M 57, 8.12 mg% controls). Changes in the control of cytoplasmic pH in stimulated PMN 0.64) and not yet receiving renal replacement therapy were may influence PMN function, and our observations may be relevant to studied. There were nine males and five females with a median the susceptibility of uremic patients to infection. age of 58 years (range 23 to 78 years). Patients with underlying disease known to be associated with abnormal PMN function, such as diabetes mellitus or systemic lupus erythematosus, Infection remains a common cause of morbidity and mortality were excluded. All patients had given informed consent and in patients with end-stage renal failure (ESRF) [1, 21. The none had received antibiotics or blood transfusion in the month pathogenesis of this susceptibility to infection is likely to be prior to study. Eleven patients established on continuous ambulatory pencomplex with abnormalities described in a variety of host defence mechanisms [3—71, but pyogenic infection is common in toneal dialysis (CAPD) for two months or longer were studied uremia providing in vivo evidence of defective phagocytic cell (mean creatinine 934 iM 91, 10.55 mg% 1.03). There were function. Consequently, the study of phagocyte function is six males and five females with a median age of 63 years (range relevant to the development of an understanding of uremia. 33 to 75 years). All patients were uninfected at the time of study Previous studies have demonstrated that the ability of circulat- and none had received antibiotics or blood transfusion in the ing neutrophils (PMN) to kill ingested micro-organisms is month prior to study. All patients had given informed consent depressed in end-stage renal failure [8, 9]. This defect is and as above, patients with underlying disease associated with reproduced in normal PMN by incubation in used peritoneal abnormal PMN function were excluded. Control cells were obtained from healthy volunteers. dialysis fluid (PDE) from patients receiving continuous ambulatory peritoneal dialysis [9]. Dialysable toxins accumulating in uremia may therefore interact adversely with circulating PMN. Peritoneal dialysis effluent Normal intracellular killing by PMN requires the interaction of oxide generation after FMLP stimulation (P < 0.05 compared to

Received for publication January 22, 1991 and in revised form April 8, 1992 Accepted for publication April 24, 1992

© 1992 by the International Society of Nephrology

Effluent peritoneal dialysate (PDE) was obtained after four hour dwells of 1.36% dialysis solution (Dianeal, Baxter Traveno! Inc. Chicago, Illinois, USA). All PDE studied were from patients established on CAPD for greater than two months who

had not received antibiotics in the month prior to study. All PDE were filtered through a 0.2 M pore filter and used fresh. 690

Haynes et at: pH changes in uremic neutrophils

691

Isolation of cells the recovery from acid loading reflects the activity of the Whole blood was anticoagulated in EDTA (final concentra- plasma membrane Na:H antiport hence is absent in sodium free tion 3.01 mM). Neutrophils were isolated by Dextran sedimen- media. The rate of recovery from acid loading in bicarbonate

tation and centrifugation over sodium metrizoate-Ficoll as free conditions thus provides a measure of the variation in previously described [12]. Isolated cells were washed twice, Na:H antiport kinetics with intracellular pH. resuspended in Hanks balanced salt solution (HBSS) with 20 Intracellular calcium measurement mM Hepes (pH 7.4) and counted in a haemocytometer. Cell viability was confirmed to be greater than 99% by Trypan blue Changes in intracellular free calcium were measured using exclusion.

Measurement of cytoplasmic pH Cytoplasmie pH was measured using the fluorescent probe 2' ,7' bis(carboxyethyl)-5(and -6) carboxyfluorescein (BCECF, Calbiochem, California, USA) essentially as described by Grinstein and Furuya [13] and Faucher and Naccache [14]. Briefly, PMN suspensions in HBSS with 20 mrt Hepes (pH 7.4, 1 X io cells/ml) were incubated with 5 LM acetylmethoxy BCECF for

the fluorescent probe Fura 2 as previously described [20, 211. Briefly, PMN were suspended at 1 x 107/ml in HBSS with 20 mM Hepes and incubated with 1 M acetylmethoxy Fura 2 for

30 minutes at 37°C with end over end rotation. Cells were

washed free of extracellular probe, resuspended at 2 x 106/ml and equilibrated for a further 10 minutes at 37°C. Loaded cells were transferred to the thermostatted cuvette of a fluorimeter (Kontron Instruments, Zurich, Switzerland). Fluorescence signals were monitored with an excitation wavelength of 340 nm five minutes at 37°C with end over end rotation. Cells were then and emission recorded at 510 nm. The stimulus FMLP was peileted by centrifugation, resuspended and incubated for a added after a stable baseline was obtained. Changes in fluoresfurther 15 minutes at room temperature. For pH measurement, cence were calibrated at the end of each experiment as previcells were diluted to 2 x 106/ml in HBSS with 20 mrvi Hepes and ously described [22]. placed in a 37°C thermostatted cuvette in a fluorimeter (Kontron Instruments, Zurich, Switzerland). Fluorescence was reSuperoxide measurement corded as the ratio of the emission recorded at 530 nm for pH Superoxide production was monitored by measuring lucigesensitive (500 nm) and pH insensitive (440 nm) excitation wavelengths. The stimulus FMLP was added after a stable nm enhanced chemiluminescence production in a LKB Wallac baseline reading was reached. The fluorescent signal was cali- 1251 Luminometer as previously described [23]. Briefly, PMN brated by exposing the cells to nigericin in the presence of were suspended in phosphate buffered saline (PBS) containing buffer solutions of known pH as previously described [15, 16]. calcium (1.0 mM/liter), magnesium (0.7 mM/liter) and 0.1% Results are expressed with the baseline pH normalized to zero. bovine serum albumin. Cells were placed in a thermostatted compartment with lucigenin (final concentration 25 sM). The stimulus FMLP was added and the superoxide production Intracellular acidification, buffering capacity and recovery monitored by measuring the total light emitted over the specifrom acid loading fied time and subtracting that of unstimulated control cells. Cells were loaded with BCECF as above but resuspended in sodium-free choline chloride buffer, and instead of stimuli 1 tM Chemicals nigericin was added to acid load the cytoplasm. A variety of intracellular pH levels were obtained by adding defatted bovine BCECF (Calbiochem) was stored aliquotted as a 1 m stock serum albumin (final concentration 5 mg/mi) to scavenge the solution in DMSO under an inert atmosphere and dessicated at nigericin and halt acidification [17]. Each aliquot of cells was —20°C. then divided into three and the cells peiieted in a microfuge. Fura 2 (Calbiochem) was aliquotted as a 1 m stock solution One aliquot was resuspended in 50 sl HBSS containing Hepes in DMSO, sonicated and stored under an inert atmosphere (pH 7.4) with the nominal absence of bicarbonate, and the initial dessicated at —20°C. Each new batch of acetylmethoxy Fura 2 linear recovery from acidification recorded over 30 seconds by was tested for the presence of partially de-esterified forms as monitoring BCECF fluorescence as described above. A second described by Scanlon, Williams and Fay [24]. aliquot was resuspended in HBSS containing 4.2 mM/liter The amiloride analogue 5-(N,N-hexamethylene)amiloride sodium bicarbonate and the recovery from acid loading re- (supplied by Dr. A. Knox, Respiratory Medicine Unit, City corded as described above. The remaining cells were resus- Hospital, Nottingham, UK) was dissolved in DMSO and diluted pended in 50 d sodium-free choline chloride buffer to deter- in aqueous buffer immediately before use. mine the intracellular buffering capacity at the same starting pH All other chemicals were purchased from Sigma Chemical as the measurements for recovery from acid loading [18]. This Company (Poole, Dorset, UK). was achieved by recording the BCECF fluorescence as deAll chemicals dissolved in DMSO were used in final dilutions scribed above after adding aS m pulse of ammonium chloride. which ensured a final concentration of DMSO less than 0.05%. The buffering capacity was calculated as previously described by measuring the increase in cytoplasmic pH above baseline in Buffers the presence of ammonium chloride [191. The rate of recovery from acid loading was calculated from the product of the The following buffer solutions were used: Hanks balanced salt solution (all mM/liter). NaC1 136, KC1 buffering capacity and the increase in pH over the first 30 seconds after acid loading, both being measured from the same 5.4, CaC12 1.3, MgSO4 0.8, KH2PO4 0.4, Dextrose 5.6 and starting cytoplasmic pH [18]. In bicarbonate free conditions, Na2HPO4 0.4 with 20 mM Hepes (pH 7.4).

Haynes et al: pH changes in urernic neutrophils

692

Table 1. Demographic data for uremic subjects

Table 2. Basal cytoplasmic pH values

Uremic patients

ESRF

Creatinine /.LM/ljter

Sodium mM/liter Potassium mM/liter Calcium mM/liter Phosphate mM/liter Albumin gluIer Bicarbonate mM/liter

Alkaline phosphatase IU Haemoglobing/dl

(N 14) 719 57

Subjects

CAPD (]V= 11)

934 91

138.2

1.4

138.5

0.9

4.4

0.2

2.1 2.1

0.1 0.1 1.0

0.1 0.1

1.5

4.4 2.4 2.0 34.8 21.0

0.1 1.9 2.4

0.5

228 9.1

0.7

38.7 16.0

260 45 8.3

39

Control ESRF CAPD

Control + PDE

Basal pH 6.82

7.00 6.91 6.81

0.068 0.091 0.060 0.054

N (22) (14) (11) (22)

Data are presented as the mean SEM of the number of subjects in parentheses.

similar abnormalities with a reduced degree of initial acidification (P < 0.05 compared to controls and P < 0.05 compared to ESRF at 20 seconds after stimulation, Fig. la). These results

were not due to a failure of the BCECF detection system to register acidic pH values in uremic cells since the addition of NaCI 132, KC1 5.4, CaC12 1.3, MgSO4 0.8, KH2PO4 0.4, the protonophore nigericin at the end of all experiments reHanks balanced salt solution with bicarbonate (all mM/liter).

Dextrose 5.6, NaHCO3 4.2 and Na2HPO4 0.4 with 20 mi Hepes sulted in detectable cytoplasmic acidification. The results for a given patient were consistent when tested several weeks later (pH 7.4). pH calibration solution (all mM/liter). KC1 120 with 20 mM (coefficient of variation less than 10%). The abnormal pH response seen in uremic PMN was reproHepes and the pH adjusted to the desired level by adding dilute duced in normal control cells by a short preincubation for five HC1 or NaOH. Choline buffer (all mM/liter). Prepared as above but with minutes at 37°C in a 1:2 dilution of uninfected four-hour dwell dialysis effluent drained from CAPD patients (PDE). Incubation isosmotic replacement of NaC1 by choline chloride. Phosphate buffered saline (all mM/liter). NaCI 137, Na2HPO4 in PDE had no significant effect upon the resting pH of cells (Table 2). Such PDE treated cells demonstrated absent cyto8.0, KH2PO4 1.47 and KCI 2.68 (pH 7.4). All buffers were prepared fresh and their pH checked prior to plasmic acidification and enhanced alkalinization after FMLP stimulation (P < 0.001 at 30 seconds and P < 0.01 at 140 use. seconds after FMLP stimulation compared to controls, Fig. lb). Statistics Using PMN from a single donor and dilutions of a single PDE, Standard error of the mean was used as a measure of variance a dose response relationship was obtained for the effect of PDE and statistical comparisons were made using the Mann Whitney treatment upon FMLP-induced cytoplasmic pH changes. Neu-

U test. Results The demographic data for the uremic patients are included in

trophil incubation in PDE as described did not decrease cell

viability assessed by Trypan blue exclusion. The pH and

osmolality of buffer containing PDE was not significantly different from that of buffer alone. Cytoplasmic pH changes in Table 1. The two groups studied were not different in terms of normal PMN were unaffected by incubation in a similar dilution age and sex distribution, hemoglobin, medication or the under- of unused, sterile peritoneal dialysis fluid (pH corrected to 7.4 lying cause of their renal disease. The CAPD group had a using dilute NaOH).

significantly elevated steady-state creatinine compared to ESRF patients (P < 0.05).

Stimulated superoxide production and intracellular calcium release in uremic PMN The acidification phase of the cytoplasmic pH response in FMLP treated PMN has been associated with respiratory burst

Cytoplasmic pH responses of stimulated uremic PMN The resting cytoplasmic pH of cells from ESRF and CAPD patients was not significantly different from that of control cells activity; the two display temporal correlation and both are (Table 2). These values are in agreement with those published absent in PMN from patients with chronic granulomatous by other authors using the BCECF and other techniques [25, disease which lack functional NADPH oxidase activity [281. A 26]. Normal cells stimulated with the chemotactic peptide defective respiratory burst may therefore result in abnormaliFMLP (1O_6 M) demonstrated a biphasic cytoplasmic pH re- ties of both bacterial killing and cytoplasmic pH in stimulated sponse; an initial rapid acidification of approximately 0.05 pH PMN. We examined respiratory burst activity in PMN from units was followed by a sustained alkalinization to approxi- uremic patients. mately 0.15 pH units above baseline by 140 seconds after The stimulated level of superoxide production was consisstimulation (Fig. la). These changes are in agreement with tently enhanced in ESRF and CAPD cells compared to controls (P < 0.05, Fig. 2). Preincubation of normal PMN in PDE as those reported by other authors [25—28]. Cells from ESRF patients demonstrated a markedly abnormal described above was associated with enhanced superoxide pH response after FMLP stimulation (Fig. La). The initial production. The PDE dilution had no effect upon cell viability acidification was consistently absent (P < 0.001 for 30 second or the buffer as described above. Superoxide production in time point compared to controls) and the degree of alkaliniza- normal PMN was unaffected by incubation in a similar dilution tion at 140 seconds after stimulation enhanced (P < 0.05 of unused, sterile peritoneal dialysis fluid (pH corrected to 7.4 compared to control), Cells from CAPD patients demonstrated using dilute NaOH).

693

Haynes et al: pH changes in uremic neutrophils 0.3

B

A

0.20

0.2

a C

a0) a

a C

E

.0

a 0)

a

0

I0.

0.10

E 0

0.1

I0.

C

C a 0) C a

a 0) C a

0

0

0.00

0.0

-0.1

—0.10

50

100

150

Time after adding FMLP, seconds

200

50

100

150

200

Time after adding FMLP, seconds

Fig. 1. The abnormal cytoplasmic pH changes stimulated by I si FMLP in (A) PMN from ESRF and CAPD patients and (B) PDE treated normal PMN compared to controls. Symbols in A are: (•) control N = 24; (0) ESRF N = 14; (D) CAPD N = 11. Symbols in B are: (•) control N = 24; (0) +PDE N = 22. Results are presented with the basal pH normalized to zero and expressed as the mean SEM. Preincubation of normal control PMN in unused peritoneal dialysis fluid had no effect upon their pH response. P < 0.05 for uremic cells compared to controls at time points between 20 and 40 seconds and at 140 seconds; P < 0.05 for CAPD cells compared to ESRF at 20 seconds after stimulation.

Other authors have related the cytoplasmic acidification seen unused, sterile peritoneal dialysis fluid (pH corrected to 7.4 after FMLP stimulation to increases in intracellular calcium; using dilute NaOH). the two are temporally correlated and abolition of the increase Sodium hydrogen antiport activity in uremic PMN in calcium levels using intracellular chelators results in loss of The abnormal pH response seen in uremic PMN may result cytoplasmic acidification [261. We therefore examined intracelfrom masking the normal acidification by enhanced alkalinizalular calcium release in PMN from uremic patients. Basal levels of intracellular calcium were similar in control tion rather than absence of a factor which normally produces and uremic PMN. The magnitude and kinetics of the calcium acidification. The alkalinization phase of the FMLP stimulated spike after FMLP stimulation were not significantly different in changes in cytoplasmic pH is sodium dependent since it is control and uremic PMN. Normal PMN preincubated in PDE as abolished under sodium free conditions, and is normally assodescribed above demonstrated a normal calcium spike although ciated with activity of the plasma membrane Na:H antiport and a trend towards increased resting calcium levels was seen(P > hence inhibited by amiloride or its analogues [25]. Hexameth0.05, not significant, Fig. 3). Calcium measurements in normal ylene amiloride (HMA) is a potent and specific Na:H antiport PMN were unaffected by incubation in a similar dilution of inhibitor (K1 0.16 tiM) which has been shown to completely

694

Haynes et a!: pH changes in uremic neutrophils

1000

30

800

Co Co0 C

EoC

Co

E

0E

600

C.)

Co

U CD

U-

Co

C,)

Co

C

400

200

C

V

U..

Co

(I)

w

w

a aFig. 2. Enhanced FMLP stimulated superoxide production by PMN from uremic patients and PDE-treated normal PMN. Each observation was recorded in duplicate and is presented as the mean SEM of the number of subjects in parentheses. Preincubation of normal control

PMN in unused peritoneal dialysis fluid had no effect upon their superoxide responses. P < 0.05 for uremic cells compared to controls.

0 Control

ESRF

CAPD

Con+PDE

Fig. 3. Normal FMLP stimulated intracellular calcium release by PMN from uremic patients and PDE treated normal PMN. Symbols

are: (Lii) basal Ca; () Ca post-stimulation by FMLP. Data are presented as the mean SEM of the number of subjects in parentheses. The kinetics of the calcium spike in uremic cells were normal. Preincubation of normal control PMN in unused peritoneal dialysis fluid had

no effect upon their resting calcium levels or the calcium spike generated by I /LM FMLP.

inhibit neutrophil antiport activity at micromolar concentrations [29], Normal PMN preincubated with HMA for five minutes at 37°C demonstrated an abnormal cytoplasmic pH creased values in uremic compared to control PMN at cytoplasresponse after FMLP stimulation (Fig. 4). The initial transient mic pH values of 6.8 or above (Fig. 5). The kinetic curve acidification was replaced by a progressive acidification and the demonstrating the relationship between Na/H antiport activity alkalinization phase was absent. Uremic PMN when similarly and cytoplasmic pH was generated by plotting the product of treated displayed progressive acidification, although to a lesser buffering capacity and the change in intracellular pH over the extent than in normal cells. Normal cells incubated in a 1:2 first 30 seconds after acid loading against the starting cytoplasdilution of PDE for five minutes and then incubated with HMA mic pH [18]. It can be seen from Figure 6 that the curves for for five minutes before stimulation with FMLP also demon- normal and uremic PMN were not significantly different, and strated progressive cytoplasmic acidification with loss of the both were concave with activity increasing as cytoplasmic pH progressive alkalinization phase seen after incubation with PDE alone (Fig. 4). Further study of alkalinization mechanisms in uremic PMN

fell. Since the buffering capacity of uremic PMN was increased, given the normal Na/H antiport kinetics, the measured change in initial cytoplasmic pH recovery after acid loading was lower

was performed by analysis of the rate of recovery from an than that seen in control PMN. In the absence of extracellular imposed acid load. Measurement of buffering capacity under sodium no recovery from acid loading occurred in uremic cells, bicarbonate free conditions demonstrated significantly in- indicating a sodium dependent mechanism (Fig. 7).

695

Haynes et a!: pH changes in uremic neutrophils 0.4

0.2

I a)

0.0

0) (0 0)

C 0)

C)

C

(0

0

—0.2

—0.4

Fig. 4. Effect of the amiloride analogue HMA (JO tM) upon the FMLP stimulated pH response of normal, PDE treated normal and ESRF PMN. Data are presented as the mean SEM of the number of at least 4 subjects for

—0.6

0

50

100

200

150

each group. Symbols are: (I) control; (•) control + HMA; (U), control + PDE; (X) control + PDE + HMA; (A) ESRF + HMA; (X) ESRF.

Time, seconds

0.034 compared to 0.239 0.038 pH U/30 seconds respectively,

N=

20,

P > 0.05 compared each pair with or without

bicarbonate). In contrast, uremic PMN demonstrated greater pH recovery over 30 seconds in the presence of bicarbonate (0.205 0.023 compared to 0.121 0.013, N = 30, P < 0.05 comparing each pair with or without bicarbonate). 0)

0a-

Discussion

C)

Uremic patients are susceptible to pyogenic infection, but the mechanisms which predispose them to this remain uncertain. Since phagocytic cells play an important role in the normal host defence mechanisms which protect against bacterial infection, it has been suggested that the function of these cells is defective in uremia. The neutrophil is the most commonly studied phagocytic cell in uremic patients, but unfortunately the literature concerning the integrity of a variety of PMN functions contains

C 0)

6.6—6.8

6.8 17.0

>7.0

Intracellular pH Fig.

5. Intracellular buffering power of control () and ESRF (El)

neutrophils at dft'erent intracellular pH levels. The buffering power was evaluated by exposing the cells to 5 mr'i NH4CI and expressed in fliM

many conflicting reports [reviewed in 4]. Some of this confusion

is generated by the heterogeneous nature of the populations studied; PMN are activated to a varying extent by contact with different dialysis membranes which will complicate any study in

H/pH unit. Results from 39 observations on control cells and 56 hemodialysis patients, and factors such as drug therapy or the SEM. The underlying disorder causing the renal failure may influence observations on ESRF cells are presented as mean buffering power of ESRF cells is significantly greater than that of PMN function. The use of different methodologies to study a controls at values of intracellular pH above 6.8 (P < 0.05 compared to given PMN function may result in further confusion [30]. For controls). example, the intracellular killing of ingested bacteria has been

reported to be normal or reduced in uremic PMN [31, 321. Accurate measurement of this PMN function requires a method Comparison of the extent of pH recovery over the initial 30 which is independent of the rate of phagocytosis, and since seconds after acid loading in paired PMN samples at the same different bacterial species display variable sensitivity to PMN starting pH demonstrated no difference in the presence or killing mechanisms, comparison between workers using different test organisms may not be valid. Nonetheless, methods absence of bicarbonate for normal PMN (mean increase 0.262

696

Haynes èt al: pH changes in urernic neutrophils

.

50

40

30 (0

0

I E

0 20

0 0 10

SS

0•0

Fig. 6. Activation of the Na:H antiport by intracellular pH in ESRF (0) and control (•) PMN. The initial velocity of recovery from acid loading was obtained for each starting pH from the product of the pH change over 30 seconds and the intracellular buffering capacity measured at the same starting pH on the same sample of cells. Results are presented for 39 observations on control cells and 56 observations on ESRF cells, and are

c0

0•

expressed in m H/liter cells/30 seconds.

0 6.2

6.4

6.6

6.8

7.0

pH

7.2

7.4

Both curves demonstrate concavity in keeping with the reported properties of the antiport, but ESRF cells behaved no differently from control cells.

which meet these technical considerations have reported de- parts of the signal transduction mechanism are intact in these creased intracellular killing of bacteria in uremic PMN and cells. The extracellular superoxide release stimulated by FMLP peritoneal macrophages from CAPD patients [9, 331. Such a defect if present in vivo would be relevant to infection in uremic was enhanced in PMN from both ESRF and CAPD patients. patients. Neutrophil killing requires the integrity of a number of The literature contains many conflicting reports concerning the cell functions such as adhesion, chemotaxis, phagocytosis and integrity of the respiratory burst hence superoxide release in degranulation [101. In turn, these functions require normal uremic PMN and many of the criticisms mentioned above are changes in cell biochemistry and well characterized techniques also pertinent. For example, decreased levels of luminol enare described to measure these events in vitro. The changes in hanced chemiluminescence have been reported in response to a intracellular calcium, superoxide release and cytoplasmic pH variety of stimuli in uremic PMN and this has been presented as which accompany activation in normal PMN are therefore well evidence of impaired respiratory burst activity in these cells described, and we have investigated the integrity of these [37, 38]. Unfortunately, luminol chemiluminescence depends in responses in uremic PMN. part upon myeloperoxidase released by primary degranulation, The resting level of intracellular calcium was not significantly hence reduced signals using this method may be the result of different from normal in PMN from ESRF or CAPD patients. either decreased respiratory burst activity, impaired degranuSimilarly, the magnitude and kinetics of the rise in calcium lation or both. Direct measurement of superoxide release using accompanying stimulation with the chemoattractant FMLP an oxygen electrode is a more sensitive technique for detecting were not different from normal in ESRF or CAPD patients. The values for the control PMN were similar to those published by many other authors using FMLP stimulated Fura 2 loaded PMN 134, 35]. We are unaware of previous reports in the literature concerning the integrity of this response in PMN from uremic

the PMN respiratory burst and a recent report using this

method has confirmed the enhanced respiratory burst activity of PMN from ESRF patients [391. This observation indicates that the supply of substrate for the oxidative killing mechanisms of uremic PMN is intact. patients. These results suggest that those enzymes and intraThe sequence of cytoplasmic pH changes which accompany cellular events activated by a rise in intracellular calcium should FMLP stimulation was abnormal in PMN from uremic patients; be intact in PMN from uremic patients. The release of intracel- the initial acidification was reduced and consequently followed lular calcium by FMLP requires normal receptor expression, by increased alkalinization. This could be the result of differactivity of a G protein and phospholipase C activity to generate ences in the buffering capacity of uremic compared to normal inositol triphosphate [361. The integrity of the calcium response PMN, absence from uremic PMN of factors responsible for to FMLP in uremic PMN may therefore indicate that these generating acidification, enhancement in uremic PMN of factors

Haynes et al: pH changes in uremic neutrophils

and uremic PMN by the concave activation curve, however, no difference was present in the cytoplasmic pH-dependent activity of the antiport in uremic and control PMN. The abnormal pH response of uremic PMN after FMLP stimulation cannot be

'I,

explained by altered Na:H antiport kinetics, although alterations in antiport expression in uremic PMN cannot be ex-

0.2

cluded by our data. Inhibition of antiport activity by the specific inhibitor HMA failed to remove the difference in cytoplasmic pH after FMLP stimulation in uremic and control PMN. Mechanisms other than the antiport may therefore be responsible for the abnormal pH response seen in uremic cells. Sodium depen-

0.0 BSA added

=0.

—0.2

dent and independent anion exchangers have also been described in the PMN plasma membrane, and under certain conditions these may contribute to cytoplasmic alkalinization [42]. The recovery from acid loading in uremic PMN was sodium dependent and, unlike control PMN, was enhanced in the presence of bicarbonate suggesting the importance of a sodium and bicarbonate transporter. Further study of anion

a)

C a)

—0.4

C a)

a) C

697

—0.6

a) C-)

exchange mechanisms in uremic PMN is indicated. Cytoplasmic pH can influence many biochemical reactions

—0.8

relevant to PMN function. For example, the activity of the enzyme 5 lipoxygenase is pH sensitive and the release by

—1.0

stimulated PMN of one of its products, leukotriene B4, a potent

mediator which enhances the respiratory burst,is dependent —1.2

20

40

60

80

Time, seconds Fig. 7. Recovery from acid loading in ESRF neutrophils is sodium

dependent. ESRF cells were incubated in HBSS in which NaCI was isosmotically replaced by choline chloride. After addition of I sM nigericin at time zero, cytoplasmic acidification was observed. After

scavenging the nigericin with 5 mg/mI bovine serum albumin, no recovery from acid loading was observed. Data are presented as mean SEM of four ESRF patients.

responsible for generating alkalinization or a combination of these factors. The cytoplasmic buffering capacity of uremic PMN was greater than that of control PMN at pH values of 6.8 or above, and this may protect the patients from acid loading. The factors responsible for the initial acidification seen in normal PMN remain controversial. The absence of this phase from the PMN of patients with chronic granulomatous disease, a condition in which the NADPH oxidase enzyme responsible for the respiratory burst is not functional, has been cited as

evidence that NADPH oxidase activity contributes to the acidification phase [28]. This evidence has been refuted by other authors who suggest instead that the rise in intracellular calcium accompanying stimulation is responsible for cytoplasmic acidification [261. In uremic PMN we have demonstrated that both the respiratory burst and calcium spike are intact after FMLP stimulation, hence absence of these factors is unlikely to explain their abnormal pH response. Our data cannot exclude the absence from uremic PMN of other factors which contribute to the acidification in normal cells. The cytoplasmic alkalinization which follows stimulation in normal PMN is largely due to the activity of a plasma membrane Na:H antiporter since it is abolished by amiloride or removal of extracellular sodium. Antiport activity is regulated by cytoplasmic pH in a variety of tissues 140, 411. This relationship was confirmed in both control

upon cytoplasmic alkalinization [29]. We are unaware of previ-

ous reports examining the cytoplasmic pH changes in stimulated uremic PMN. The abnormal pH response of uremic PMN

may relate to functional defects, and further study of its relationship to degranulation and the miroenvironment of the phagocytic vacuole are relevant to intracellular killing by uremic PMN. We have previously demonstrated the presence of defective killing in PMN from CAPD patients, and the appearance of factors which impair bacterial killing by normal PMN in PDE during a four-hour dwell time [9]. In these experiments we have demonstrated that short incubations in four-hour dwell PDE produce biochemical defects in normal PMN consistent with those present in uremic cells. The enhancement of superoxide release by factors in PDE has previously been reported, but we are unaware of publications concerning its effect upon the calcium and pH responses of PMN. These observations are relevant to the pathogenesis of peritonitis in CAPD patients where the intracellular survival of bacteria within PMN present in infected PDE has been clearly demonstrated [43]. Controling experiments using PDE is difficult; it is not possible to obtain fluid from the peritoneum of non-anesthetised normal controls, and interspecies variation in PMN physiology can make animal models difficult to interpret. In our experiments care was taken to exclude nonspecific effects of the pH or osmolality of PDE upon PMN, and the controls with unused dialysis fluid indicate that the factors influencing PMN function require interaction between fluid and the peritoneum. At the short incubation times employed in these experiments, the effects of PDE are unlikely to be due to endotoxin contamination. Preliminary characterization of the factors in PDE which influence PMN function indicate the involvement of a non-protein molecule of a molecular weight less than 1500 Da; further characterization is in progress. Characterization is required before the factor in PDE producing defective PMN functions can be identified as the circulating toxin impairing circulating PMN function in uremia.

698

Haynes et a!: pH changes in uremic neutrophils

Abnormal PMN function may be relevant to the pathogenesis of infection in uremic patients. This study indicates the need for further investigation of the influence of uremia upon these cells.

Moreover, since the biochemistry and physiology of normal PMN is well described, abnormalities in uremic PMN may provide insight into the mechanism of dysfunction in other uremic tissues. Acknowledgments These studies were supported in part by grants from the National Kidney Research Fund, Great Britain and Baxter Healthcare Corporation, Deerfield, Illinois, USA. We express our thanks to the staff of the CAPD unit for their assistance.

Reprint requests to John Fletcher, M.D., Department of Haematology, City Hospital, Hucknall Road, Nottingham, NG5 JPB, England, United Kingdom.

Appendix. Abbreviations

BCECF, 2,7-bis(2 carboxyethyl)-5 ,(6)-carboxyfluorescein CAPD, continuous ambulatory peritoneal dialysis DMSO, dimethylsulphoxide EDTA, ethylenediaminetetra-acetic acid ESRF, end-stage renal failure FMLP, formylmethionyl leucyiphenylalanine HBSS, Hank's balanced salt solution HMA, hexamethylene amioride PBS, phosphate buffered saline PDE, effluent peritoneal dialysis fluid PMN, polymorphonuclear neutrophil

14. FAUCHER N, NACCACHE PH: Relationship between pH, sodium and shape change in chemotactic factor stimulated human polymorphonuclear neutrophils. J Cell Physiol 132:483—490, 1987 15. THOMAS JA, BUCIiSBAUM RN, ZIMNIAK A, RACKER E: Intracellu-

lar pH measurements in Ehrlich ascites tumour cells utilising spectroscopic probes generated in situ. Biochem 18:2210—2218, 1979 16. SULLIVAN R, GRIFFIN JD, WRIGHT J, MELNICK DA, HORNE JM, LEAVITT JL, FREDETFE JP, LYMAN CA, LAZZARI KG, SIM0N5 E:

Effects of rh GM CSF on intracellular pH in mature neutrophils. Blood 721:1665—1670, 1988 17. GRIN5TEIN S, FURUYA W: Characterization of the amiloride sensi-

tive Na:H antiport of human neutrophils. Am J Physiol 250:283— 291, 1986 18. ABDULKARIM M SALEH, BATTLE DC: Kinetic properties of the Na:H antiporter of lymphocytes from the spontaneously hypertensive rat: Role of intracellular pH. J Clin Invest 85:1734—1739, 1990 19. Roos A, BORON WF: Intracellular pH. Physiol Rev 61:296—434, 1981

20. AL MOHANNA F, HALLET M: The use of Fura 2 to determine the relationship between cytoplasmic free calcium and oxidase activation in rat polymorphonuclear neutrophils. Cell Calcium 9:11—15, 1988

21. GRYNKIEWICZ G, POENIE M, TSIEN RY: A new generation of calcium indicators with greatly improved fluorescent properties. J Biol Chem 260:3440—3450, 1985

22. TSIEN RY, POZZAN T, RINK TJ: Calcium homeostasis in intact

lymphocytes: Cytoplasmic free calcium monitored with a new intracellularly trapped fluorescent indicator. J Cell Biol 94:3325— 3334, 1982 23. GYLLENHAMMAR H: Lucigenin chemiluminescence in the assess-

ment of neutrophil superoxide anion production. J immuno! Method 97:209—214, 1987 24. SCANLON M, WILLIAMS DA, FAY FS: A calcium insensitive form

of Fura 2 associated with polymorphonuclear leucocytes. J Biol Chem 262:6308—6313, 1987

25. WEISSMAN S, PUNZO A, FORD C, SHA'AFI RI: Intracellular pH

changes during neutrophil activation: Na:H antiport. J Leu Biol

References 1. KEANE WF, Rxn LR: Host defenses and infectious complications in maintenance haemodialysis patients, in Replacement Of Renal Function By Dialysis (2nd ed), edited by DRUKKER W, PARSON FM, MAHER iF, Boston, Martinus Nijhoff, 1983, pp 646—648 2, HIGGINs RM: Infections in a renal unit. Quart J Med 70:41—49, 1989

3. GOLDBLUM SE, REED WP: Host defenses and immunological alterations associated with chronic haemodialysis. Ann in! Med 93:587—613, 1980

4. Lewis SL, VAN Es D: Neutrophil and monocyte alterations in chronic dialysis patients. Am J Kid Dis IX(5):381—395, 1987 5. Luccn L, CAPPELLI G, ACERB! MA, SPATTINI A, LUSVARGHI A:

41:25—32, 1987

26. NACCACHE PH, THERRIEN S, CAON AC, LIA0 N, GILBERT C, McCou SR: Chemoattractant induced cytoplasmic pH changes and cytoskeletal reorganisation in human neutrophils. J Imrnunol 142: 2438—2445, 1989 27. NACCACHE PH, FAUCHER N, BORGEAT P. GASSON J, DIPERsI0 iF:

Granulocyte-macrophage colony stimulating factor modulates the

excitation responses coupling sequence in human neutrophils. J immunol 140:3541—3548, 1988 28. GRINSTEIN 5, FURUYA W, BIGGAR W: Cytoplasmic pH regulation in normal and abnormal neutrophils. J Biol Chem 261:512—517, 1986 29. OsAKI M, SUMIM0T0 H, TAKASHIGE K, CRAGOE EJ, H0RI Y, MINAKERRI 5: Na:H exchange modulates the production of LTB4 by human neutrophils. Biochem J 257:751—758, 1989

Oxidative metabolism of polymorphonuclear leucocytes and serum opsonic activity in chronic renal failure. Nephron 51:44—SO, 1989

30. HAYNES AP, FLETCHER J: Neutrophil function tests. Balliere's

6. SCHOLLMEYER P, BOZKHART F: The immune status of the uraemic

The neutrophil function of patients treated by haemodialysis or

patient: haemodialysis v continuous peritoneal dialysis. Clin Nephrol 30(Suppl 1):S37—540, 1988

7. TOLKOFF RUBIN NE, RUBIN RH: Uraemic and host defenses. N Engi J Med 322(1 1):770—772, 1990 8. BUGGY BP, SCHABERG DA, SCHWARTZ RD: Intraleucocytic se-

questration as a cause of persistent Staph. aureus peritonitis in CAPD. Am J Med 76:1035—1040, 1984 9. HARVEY DM, SHEPPARD K, MORGAN AG, FLETCHER J: Neutrophil

function in patients on continuous ambulatory peritoneal dialysis. Br J Haem 68:278—284, 1987

10. LEHRER RI, GANZ T, BABIOR B: Neutrophils and host defence. Ann mt Med 109:127—142, 1988

11. SPITZNAGEL JK: Antibiotic proteins of human neutrophils. J C/in Invest 86:1381—1386, 1990

12. BOYUM A: Isolation of mononuclear cells and granulocytes from human blood. Scand J C/in Lab invest 2l(Suppl 97):77—79, 1968 13. GRINSTEIN 5, FURUYA W: Cytoplasmic pH regulation in phorbol ester activated human neutrophils. Am J Physiol 251 :C55—C60, 1986

Clinical Haematology 3(4):871—887, 1990 31. HUTTENEN K, LAMPAINIE E, SILVENNOINEN KS, TIILIKAINEI A:

CAPD. Scandf UrolNephrol 18:167—172, 1984 32. MCINTOSH J, HANSEN P. ZIEGLER J, PENNY R: Defective immune

and phagocytic functions in uraemia. mt Arch Allergy immunol 5l(5):544—549, 1976 33. MACGREGOR Si, BROCK JH, BRIGGS JD, JUNOR BJR: Bactericidal

activity of peritoneal macrophages from CAPD patients. Nephrol Dial Transplant 2:104—108, 1987 34. AL MOHANNA FA, HALLET MB: The use of Fura 2 to determine the

relationship between free calcium and oxidase activation in rat neutrophils. Cell Calcium 9:17—25, 1988 35. PERIANNIN A, SYNDERMAN R: Analysis of calcium homeostasis in activated neutrophils, J Biol Chem 264(2);lOOS—1009, 1989 36. ALLEN RA, TRAYNOR AE, OMAN GM, JEsAITIs AJ: The chemo-

tactic peptide receptor. Haema:ol Oncol Clin N Am 2(1):33—39, 1988 37. LUCCHI L, CAPPELLI 0, ACERB! MA, SPATTINI A, LUSVARGHI E:

Oxidative metabolism of neutrophils and serum opsonic activity in chronic renal failure. Nephron 51:44—50, 1989 38. ECKHARDT UK, ECKHARDT H, HUBER MJ, ASSCHER AW: Analysis

Haynes er al: pH changes in uremic neutrophils

of neutrophil respiratory burst activity in uraemic patients using

699

41. GRINSTEIN 5, Rorit D, MASON MJ: Na:H exchange and growth

factor induced cytosolic pH changes. Role in cell proliferation.

whole blood chemiluminescence. Nephron 43:274—278, 1986 39. PAUL JL, ROCH-ARVEILLER M, MAN NK, LUONG N, MOATTI N, RAICHVARG D: Influence of uraemia on polymorphonuclear neutro-

42. SIMcI-iowITz L: Anion exchange in human neutrophils, in Cell

phi! oxidative metabolism in end stage renal failure and dialysed

Physiology Of Blood, edited by GUNN, PARKER, Society of General

patients. Nephron 57:428—432, 1991 40. HOFFMAN ER, SIMONSEN LO: Volume and pH regulation. Physiol Rev 69:317—371, 1989

Biochim Biophys Act 988:73—97, 1989

Physiology 41st Annual Symposium, Rockefeller University Press 43. GOULD IM, CASEWELL MW: The laboratory diagnosis of peritonitis during CAPD. J Hosp Inf 7:155—160, 1986