Isolation of Modified Ubiquitin as a Neutrophil Chemotaxis

Inhibitor from Uremic Patients

GERALD COHEN, MICHAEL RUDNICKI, and WALTER H. HORLDivision of Nephrologv. Departnieizt of Medicine, Univei�s-itv of Vienna, Vienna, Austria.

Abstract. Uremic toxins are factors that accumulate in the

serum and penitoneal cavity of unemic patients. They are re-

sponsible for a variety of functional disturbances and also

contribute to the increased risk of infection by interfering with

essential functions of the unspecific immune response. From

the peritoneal effluent of penitoneal dialysis (PD) patients, a

pcptide was isolated by applying three different chromato-

graphic methods. This peptide inhibits the chemotactic move-

ruent of pobyruorphonuclear leukocytes (PMNL) in an in vitro

assay in a concentration-dependent, nonreversible manner, and

therefore belongs to the group of urcruic toxins. Amino acid

sequencing showed that the isolated peptide has the same

amino terminal sequence as ubiquitin. The peptide also reacted

with anti-ubiquitin antibodies in a Western blot experiment,

but had a more acidic isoebectnic point than ubiquitin. By using

affinity chromatography, anti-ubiquitin antibody binding frac-

tions were isolated from all PD and hemodialysis (HD) patients

investigated. These fractions contained the same acidic band

and also significantly inhibited PMNL chemotaxis. Ubiquitin

per se had no effect on PMNL cheruotaxis. Therefore, it is

concluded that from PD and HD patients a modified form of

ubiquitin was isolated, and this modification was responsible

for its inhibitory effect. (J Am Soc Nephrol 9: 451-456, 1998)

One of the major factors causing the high incidence of mor-

bidity and mortality among uremic patients undergoing hemo-

dialysis (HD) and penitoneal dialysis (PD) treatment is the

increased risk of bacterial infections (1-3). Between 15 and

20% of all cases of mortality among hemodialysis patients are

due to infection (summarized in reference 4). An impairment

of the unspecific immune defense system, especially of poly-

ruorphonuclear leukocytes (PMNL), is primarily responsible

for this situation (5). PMNL are cells of the first-line immune

defense and are also the main immune cells in the penitoneal

cavity during bacterial infections (6).

A variety of factors contribute to the diminished functions of

PMNL found in uremia. Toxic substances that directly inter-

fere with PMNL activities accumulate in the serum of uremic

patients and are partly responsible for their susceptibility to

bacterial infections. A number of urcmic toxins have already

been purified and characterized. In our laboratory, a gnanubo-

cyte inhibiting protein with homology to Ig bight chains has

been isolated (7). We could show that free Ig light chains per

se are able to interfere with essential PMNL functions (8). A

granubocyte inhibiting protein with homology to p2-micro-

globulin was also isolated from hemodialysis patients (9) and

from patients undergoing continuous ambulatory penitoneab

dialysis (CAPD) treatment ( 10). Angiogenin was purified from

uremic patients as a PMNL degranulation-inhibiting protein

( 1 1 ), and complement factor D was shown to adversely affect

Received February 10, 1997. Accepted August 1 1, 1997.Correspondence to Dr. Gerald Cohen. Medizinische Universit#{228}tsklinik III.

Wahringer GUrtel 18-20. A-l090 Vienna, Austria.

l()46-6673/0903-045 I $03.()0/0

Journal of the American Society of Nephrology

Copyright 0 1998 by the American Society of Nephrology

PMNL functions (12,13). Furthermore, p-cresol was identified

as a uremic solute that impairs the respiratory burst activity of

PMNL (14).

In the present study, we describe the isolation of a modified

version of ubiquitin from the penitoneal effluent of PD patients

and from the ultrafiltrate of high-flux HD patients. We show

that the purified protein significantly inhibits PMNL chemo-

taxis, whereas unmodified ubiquitin has no effect on this

essential PMNL function.

Materials and Methodsisolation of the Chemotaxis-inhibiting Protein

Patients. The chemotaxis-inhibiting protein (CIP) was originally

isolated from the effluent of a patient (patient I ) undergoing CAPD,

using three different chromatographic methods. The patient was

treated with 2 L of PD solution four times daily. CIP was also isolated

by affinity chromatography from the same patient undergoing nightly

intermittent PD (NIPD) treatment and from two other CAPD patients

(patients 2 and 3) treated in the same way as patient I . From theultrafiltrate of three HD patients (patients 4 to 6), CIP was isolated as

well. Informed consent was obtained from all patients. Care was taken

that the patients did not have any clinical signs or symptoms of

infection interfering with our study. The serum C-reactive protein

concentration was lower than the detection limit (0.5 rug/dI). The PD

effluents were without any laboratory findings responsible for perito-

neal infections.

isolation of CIP by Three Chromatographic Methods

High molecular weight proteins were removed from the CAPD

effluent by using a polysulfone membrane with a cutoff of approxi-

mately 60 kD (F60; Fresenius, Oberursel, Germany). Then, the CAPD

effluent was concentrated about 30-fold, and the NIPD effluent was

concentrated about 100-fold with Cuprophan membranes, with a

cutoff of approximately 5 kD (Hemoflow E3 or E4, Fresenius).

Afterward, the concentrate was chromatographed on a fast-protein

452 Journal of the American Society of Nephrology

liquid chromatographic (FPLC) system (Pharmacia LKB Biotechnol-

ogy, Uppsala, Sweden), and the protein was detected with a variable

wavelength monitor (VWM 2141, Pharmacia) at wavelengths of 225

and 214 nm.Ion Exchange Chromatography. In a first step, the concentrate

was applied to FPLC, using an ion exchange Mono Q HR 10/10

column equilibrated with 20 mM Tnis-HC1, pH 8.0. The flow rate was4 ml/min. The biological activity (see below) was found in the

unbound material and was concentrated by lyophilization. resus-pended in water, and dialyzed against 200 ruM sodium phosphate

buffer, pH 7.2, using a Spectra/Por molecular-porous membrane(Spectrum, Houston, TX) with a molecular cutoff of approximately 1

kD. The bound material that was inactive in our in vitro assay was

eluted with 1 M NaCI in 20 mM Tris-HC1, pH 8.0. to regenerate the

column for next use.

Size Exclusion Chromatography. The sample was then chro-

matographed on FPLC with a Superdex 75 prep grad column using200 mM sodium phosphate buffer, pH 7.2, and a flow rate of 1ml/min. The protein peak containing the biological activity was in the

molecular weight range of 8 to 10 kD, was concentrated by lyophi-

lization, dissolved in water, and dialyzed against water.

Reversed-Phase Chromatography. The material was furtherchromatographed on FPLC with a Resource 1 -ml reversed-phase

chromatography column (Pharmacia), using a gradient from 0 to

100% acetonitnile with 0.05% (vol/vol) trifluoroacetic acid (Aldrich

Chemical Co., Milwaukee, WI) with a corresponding gradient of

water from 100 to 0%. The flow rate was 4 mI/mm. Peak fractions

have been extensively dialyzed against phosphate-buffered saline

(PBS; pH 7.2). The fraction that had been eluted at 30% acetonitrile

showed the highest biological activity and was used for further in-vestigations.

Isolation of C/P by Immunoaffinitv Chromatography

Concentrates of the PD effluents were prepared as described above.

After identification by amino acid sequencing of CIP as being homol-ogous to ubiquitin, concentrates of the PD effluents that were prepared

as described above and ultrafiltrates obtained from HD patients were

passed through an immunoaffinity column specific for ubiquitin. For

the preparation of this column, anti-ubiquitin antiserum (U-5379,

Sigma) was coupled onto a N-hydroxysuccinimide-activated HiTrap

affinity column (Pharmacia), according to the manufacturer’s instruc-

tions. After washing the column with PBS, the bound material was

eluted by lowering the pH (0.2 M Glycine-HCI, pH 2.8). The fractionswere neutralized immediately after elution by a 1/5 vol of I M

Tris-HC1, pH 8.0. The protein-containing fractions were identified by

the bicinchoninic acid protein assay (Pierce, Rockford, IL). They werepooled, dialyzed against PBS diluted with water 1 to 40, lyophilized,

and then taken up in water in a 40th of the original volume so that the

sample was finally dissolved in I X PBS.

Characterization of C/P

Molecular Characterization. Sodium dodecyl sulfate-polyac-

rylamide gel clectrophoresis (SDS-PAGE). 8 to 25% gels, or high-density gels and isoelectric focusing (p1 range, 3 to 9) was performed

using the Phast System (Pharmacia). For the reducing SDS-PAGE,

10% (vol/vol) beta-mercaptoethanol was added to the sample buffer,

and the sample was boiled for 3 ruin before loading. The reactivity ofCIP with an anti-ubiquitin antibody was tested by Western blotting.

After electrotransfer of the proteins from the SDS polyacrybamide gelsto a nitrocellulose membrane, the transferred proteins were detectedwith an anti-ubiquitin antiserum (U-5379, Sigma), horseradish perox-

idase-labeled goat-anti-rabbit antibodies, and the enhanced chemilu-

minescence detection system (Amersharu International, Little Chab-

font, Buckinghamsire, United Kingdom). The amino acid sequence

was determined by automatic Edruan degradation (model 476A; Ap-

plied Biosystems, Foster City, CA) in the protein chemistry laboratory

of the Institute for Biochemistry of the University of Vienna by Dr. R.

Prohaska.

PMNL Isolation

PMNL were isolated as described previously ( 15). Briefly, 10 ml of

the venous whole blood of healthy donors was anticoagulated with 10 jd

of heparin sodium (Liquemine Roche iv; Hoffmann-Roche, Bascl, Swit-

zerland) and put on top of 10 ml of Ficoll-Paque (Pharmacia). After

approximately 45 mm, most of the erythrocytes sedimented into the

Ficoll layer. A Percoll (Pharmacia) step gradient (63% vol/vol and 72%

vob/vol) was used to separate granulocytes, lymphocytes, monocytes, and

thrombocytes that stayed in the plasma supernatant. The PMNL band was

isolated, the PMNL were washed twice in PBS, and the cells counted in

a cell counter (Cell Dyn 610, Abbott, Wiesbaden, Germany). The via-

bility of the PMNL obtained by this protocol was greater than 95% as

determined by the trypan blue (Sigma) exclusion assay.

Chemotaxis Assay

We used the under-agarose method ( 16) to test the inhibitory effect

of our fractions or of the purified factor on PMNL chemotaxis. PMNL

were resuspended in PBS at a concentration of I X I0� cells/lO �l in

PBS or in PBS containing the isolated peptide at a final concentrationof 20 j.�g/ml and incubated for 15 ruin at 37#{176}Cbefore starting the

chemotaxis assay. N-formyl-ruethionyl-leucyl-phenylalanine (Sigma),

dissolved in Hank’s buffer with Ca2� and Mg2� (Seromed, Bio-

chrom, Berlin, Germany), was used as a chemoattractant at a concen-

tration of2.5 X l0� M. Afterward, the agarose plates were incubated

for 100 ruin at 37#{176}C.Then the cells were fixed with methanol andparaformaldehyde and stained with Gierusa (Merck, Darmstadt, Ger-

many). The distance the cells migrated under the agarose was mea-

sured under the microscope. For each isolate, between eight and 12

independent assays were performed. The results were statistically

evaluated by paired analyses.

ResultsInitial isolation and Characterization of C/P

We concentrated the penitoncab effluent obtained from pa-

tient 1 undergoing CAPD treatment, as described in Material

and Methods. To obtain a pure substance that shows chemo-

taxis inhibiting activity, we used three different chromato-

graphic methods (see Material and Methods). After each pun-

fication step, i.e. , ion exchange chromatography, gel filtration,

and reversed-phase chromatography, we controlled the biolog-

icab activity of the individual fractions in our in vitro chemo-

taxis assay. Furthermore, we checked the purity of the most

active fractions by nonreducing SDS-PAGE. Figure 1A shows

the progress of purification using the concentrated CAPD

effluent of patient 1 , resulting in the pure protein with chemo-

taxis inhibiting activity. Reducing (Figure 1 B, lane a) and

nonreducing (Figure 1B, lane b) SDS-PAGE of CIP show a

single band, indicating that we obtained a pure pcptidc that

does not consist of subunits connected by disulfide bridges.

The sequence of the 16 N-terminal amino acids of CIP is

Mct-Gln-Ilc-Phc-Val-Lys-Thr-Leu-Thr-Gly-Lys-Thr-Ibe-Thr-

67.0

43.0

30.0

20.1

1 4.4

1 7.2014.60

8.246.38

2.56

abmabcdM ab

Modified Ubiquitin as PMNL Chemotaxis Inhibitor 453

A B C

!�Figure 1. (A) Progress of chemotaxis-inhibiting protein (CIP) purification. Nonreducing sodium dodecyl sulfate-polyacrylamide gel electro-

phoresis (SDS-PAGE) on an 8 to 25% gradient gel. Lane a, concentrated peritoneal dialysis (PD) effluent from patient I ; lanes b through d,

active fractions after ion exchange (b), size exclusion (c), and reversed-phase chromatography (d). M, high molecular weight marker. Molecular

weight is indicated in kilodabtons. (B) SDS-PAGE ofCIP on a high-density gel. Reducing (a) and nonreducing (b) conditions. m, low molecular

weight marker. Molecular weight is indicated in kibodaltons. (C) Western blotting experiment. Lane a, commercially available ubiquitin; lane

b, purified CIP.

Lcu-Glu. This is exactly the same primary structure as found at

the N terminus of ubiquitin (17).

Nonreducing SDS-PAGE of commercially available ubiq-

uitin (Figure 2. lane a) and of CIP (Figure 2, lane b) shows that

both peptides comigrate, and therefore have about the same

size. The molecular mass of ubiquitin predicted from its amino

acid sequence is 8.5 kD ( 17). The anomalous migration of

ubiquitin as a 5.5-kD band on the SDS polyacrybamide gel

indicates the incomplete unfolding of the polypcptide chain.

This finding is consistent with the highly globular compact

conformation of ubiquitin (1 8). In a Western blotting cxpcni-

ment (Figure bC), we found that CIP is able to react with

anti-ubiquitin antibodies. Isoclectric focusing shows that CIP

has an isoclectric point of 5.2 (Figure 3, lane a). This is much

more acidic than the isoelectnic point of unmodified ubiquitin

of approximately 7 (Figure 3, lane b) (19).

inhibitory Effect of C/P on PMNL Chemotaxis

Figure 4 shows that CIP significantly inhibited the cheruo-

tactic movement of PMNL from healthy donors. On the other

30.0

20.114.4 I

--- _

Mabcdefgh I

Figure 2. Nonreducing SDS-PAGE on an 8 to 25% gradient gel. M,

high molecular weight marker. Lane a, commercially available ubiq-

uitin; lane b, CIP isolated from the continuous ambulatory PD

(CAPD) effluent of patient I by three chromatographic methods: lanesc and d, CIP isolated from the nightly intermittent PD effluent of

patient I by affinity chromatography: lanes e and f, CIP isolated by

affinity chromatography from the CAPD effluent of patients 2 and 3,

respectively: lanes g through i, CIP isolated by affinity chromatogra-

phy from hemodialysis (HD) patients 4, 5, and 6, respectively.

hand, commercially available unmodified ubiquitin had no

effect. Furthermore, we showed that the inhibitory effect of

CIP was not reversible (Figure 4). Nevertheless, the viability of

PMNL was not influenced in the presence of CIP, as deter-

mined by the trypan blue exclusion assay (data not shown). We

also found that CIP inhibits PMNL chemotaxis in a concen-

tration-dcpendent manner (Table 1 ). We tested the ability of

CIP to attract normal PMNL. This was not the case within the

concentration range tested (2.5 X lO_6 M to 2.5 X b0” M)

(data not shown).

Affinity Chromatography and Characterization of Anti-

Ubiquitin Antibody Binding Fractions

Our data suggested that we isolated a modified form of

ubiquitin from the penitoneab effluent and that this modification

is responsible for the inhibiting effect on PMNL cheruotaxis. In

a second set of experiments, we prepared an affinity column

specific for ubiquitin and applied the penitoneal effluent ob-

tamed from the same patient (at this time undergoing NIPD

5.1

6.0

6.5

M abcdefgFigure 3. Isoelectnic focusing of CIP isolated from the CAPD effluent

of patient I by three chromatographic methods (a), commercially

available ubiquitin (b), and CIP isolated from patients 2 through 6 by

affinity chromatography (c through g). M, p1 marker.

120

1 00

80

% 60distancemigrated

40

20

0

120

1 00

80

% 60distance

migrated

40

20

0

a CIP, chemotaxis-inhibiting protein; PMNL, polymorphonuclear

beukocytes: fMLP, N-formyl-methionyl-leucyl-phenylalanine.

454 Journal of the American Society of Nephrology

- 1 ir U

Figure 4. Chemotactic movement toward N-formyl-ruethionyl-leucyl-

phenylalanine (fMLP) of polymorphonuclear leukocytes (PMNL)

from healthy subjects in the presence of CIP purified from patient 1

(1 ), commercially available ubiquitin (U), or after preincubation with

and subsequent removal of CIP purified from patient 1 ( 1 r). Mean ±

SEM. *P < 0.05: � < 0.01.

Table 1. Effect of different concentrations of CIP on the

chemotactic movement of PMNL from healthy

subjects toward fMLP”

�g/ml % Migration (mean ± SEM)

0 100

4 101±3

9 96±5

15 75±2

20 46±7

treatment) onto the column. Figure 2 (lanes c and d) shows the

result of an SDS-PAGE of the material bound to the anti-

ubiquitin affinity column from two different purification cx-

periments. We obtained in both cases a single band migrating

at the same distance as ubiquitin.

We further applied the concentrated CAPD effluent from two

other PD patients (patients 2 and 3) and from three HD patients

(patients 4, 5, and 6) onto the anti-ubiquitin affinity column.

SDS-PAGE of the bound material (Figure 2, lanes e through i)

showed a single band comigrating with that of ubiquitin.

Isoclectric focusing of the material ebuted from the anti-

ubiquitin column showed that the fractions isolated from the

other patients contained a component with an acidic isoclectnic

point of approximately 5.2 (Figure 3, lanes c through g) that

most likely corresponds to a modified form of ubiquitin.

Chemotaxis-/nhibiting Activity of Anti- Ubiquitin

Antibody Binding Fractions

We tested all fractions obtained by immunopurification in

our in vitro chemotaxis assay. Figure 5 shows that the samples

- U 1’ 1’’ 2 3 4 5 6

Figure 5. Chemotactic movement toward fMLP of PMNL from

healthy subjects in the presence of commercially available ubiquitin

(U) and of immunopunified CIP from patient I (1 ‘, I”), the PD patients2 and 3 (2 and 3, respectively), and from HD patients 4, 5, and 6 (4,

5, and 6, respectively). *P < 0.05; **P < 0.01.

obtained from all PD patients and from all HD patients signif-

icantly inhibited the cheruotactic movement of PMNL.

DiscussionIn the present study, we isolated from PD and HD patients a

pcptidc with homology to ubiquitin and showed that it has the

ability to inhibit the chemotactic movement of PMNL in a

concentration-dependent manner.

Uremic patients have an increased risk of infection, and as a

result they have a high incidence of morbidity and mortality

(20). In fact, despite intensive research and improved clinical

conditions within the past 20 years, bacterial infections are still

one of the leading causes of death in patients with chronic renal

failure. Because PMNL play a key role in the unspecific

immune defense against bacterial infections, the function of

PMNL isolated from uremic patients has been studied inten-

sively.

These investigations of the PMNL-ruodubating activities of

uremic toxins arc an important contribution to a better under-

standing of the immune status of patients with chronic renal

failure and to future applications in diagnosis and therapy. The

PD effluent represents an important starting material for the

isolation of uremic toxins: The fact that many disturbed func-

tions of PMNL can be corrected by PD treatment is an mdi-

cation for the presence of uremic toxins in the peritoncal cavity

of PD patients. Furthermore, it was shown that addition of

CAPD effluents to PMNL of healthy donors markedly de-

pressed phagocytosis (2 1 ) and also interfered with their ability

to kill ingested Staphylococcus epidertnidis (22).

To have a uremic toxin, the compound should be retained in

the uremic serum, and the concentration should be higher in the

ureruic than in the normal plasma. Although the majority of

proteins that appear in penitoncal effluent come from the vas-

cuban compartment, it is possible that some are generated from

Modified Ubiquitin as PMNL Chemotaxis Inhibitor

cells within the penitoneal cavity on within the peritoncab ruem-

brane. We therefore looked up this component in ureruic serum

ultrafiltrate to exclude that CIP is only extracted from perito-

neal effluent and produced in the peritoncuru. This study

provides evidence that the protein is present in the blood of

unemic patients. The adequate removal of ureruic solutes in the

higher molecular weight range during high-efficiency and

high-flux hemodiabysis has been studied (23).

An obvious consequence of the partial removal of PMNL-

inhibiting factors from the uremic serum by PD treatment and

consequently their presence in the peritoneal cavity is that these

toxins exert a negative influence on functions of peritoneab

PMNL. Whereas most of the cells within the peritoneal cavity of

a PD patient are macrophages in a stable situation, after the onset

peritonitis that is accompanied by a massive and rapid influx of

cells into the peritoncal cavity, more than 85% of the cells are

PMNL (6). Therefore, urcmic toxins affecting PMNL functions

seem to be at least partly responsible for the high susceptibility of

PD patients to peritonitis and, hence, for the increased risk of

morbidity and mortality in this group of patients.

Chemotaxis of PMNL is the first important step of the unspe-

cific immune response (24). Whereas the transendotheliab migra-

tion of PMNL is mainly regulated by interlcukin-8 (25), N-

formyb-methionyl peptides that arc generated by the invading

microorganisms are the main chcmoattractants in bacterial infec-

tions and induce PMNL finally to move to the site of infection

(26). Therefore, we used the peptide N-formyl-methionyl-leucyl-

phenylalaninc, one of the most potent within this group of che-

moattractants (27), in our in vitro chemotaxis assay.

We demonstrated that CIP significantly inhibits PMNL che-

motaxis (Figure 4). Because it has been described (28) that PD

treatment per se might induce the generation of a chcmoattrac-

tant, we tested this ability. However, CIP did not attract normal

PMNL.

Sequencing of the N terminus showed that the first 16 amino

acids of CIP arc identical to those of ubiquitin. Ubiquitin is a

polypcptide with a molecular weight of approximately 8.5 kD.

Its 76 amino acids are highly conserved in evolution. In our

control experiments, we used bovine ubiquitin that has the

same amino acid sequence as human ubiquitin (29). Ubiquitin

is present in all cells investigated so far, i.e. , in bacteria, yeast,

higher plants, and animals (30). It can be attached in an

ATP-nequining process to other proteins via an isopeptidc link-

age between its C-terminal carboxyl group and the epsilon-

amino groups of lysinc residues in the target protein. This

posttransbational protein modification has a variety of biobog-

ical effects. The most prominent function is the participation of

ubiquitin in the protcosomal protein degradation (3 1 ): Ubiq-

uitin covalently bound to proteins can target them to a complex

protcase called the 265 protcasome that degrades misfolded

proteins within the cell. Usually, the formation of ubiquitin

oligoruers is necessary for proteasome binding and subsequent

protein destruction (32). The fact that we isolated by affinity

chromatography only material that migrates as a single band in

SDS-PAGE (Figure 2) at the same distance as ubiquitin mdi-

cated that the anti-ubiquitin antibody-binding fractions con-

tamed only ubiquitin and/or a modified form of it, but no

ubiquitin oligoruers or other proteins with ubiquitin covabently

attached to them.

The increased protein degradation beading to muscle wasting

in uremia is caused by metabolic acidosis induced ATP- and

ubiquitin-dependent protcolysis (33,34). Furthermore, the

ubiquitination of short-living regulatory proteins plays an im-

pontant role in modulating their intracellular concentrations.

Besides those intracellular functions of ubiquitin, excreted

ubiquitin was isolated from the culture supernatant of oocytes

and shown to inhibit the cytotoxic properties of platelets (35).

A significantly increased level of ubiquitin was described in

the plasma of uremic patients and patients undergoing hemo-

dialysis treatment (36). At present, we cannot give exact in vivo

concentrations of the modified ubiquitin in uremic patients.

CIP has about the same size as ubiquitin as shown by

SDS-PAGE (Figure 2), but has a more acidic isoclcctnic point

of 5.2. Because ubiquitin did not show any chcmotaxis inhib-

itory activity, we conclude that CIP is a modified form of

ubiquitin and that this modification is responsible for its bio-

logical activity. Two types of modifications, i.e. , carbamoyla-

tion and glycation, could explain our findings, because both

reactions bead to a loss of positive charge by modifying lysine

residues in proteins. The molecular weight change caused by

either modification would not necessarily be big enough to be

detected by SDS-PAGE.

A small percentage (approximately 0.8%) of urea that accumu-

lates in the scra of urcmic patients is converted spontaneously to

cyanate (37). Cyanate can react with the epsilon-amino groups of

lysins, leading to a carbamoylated protein (38). Glucose can react

noncnzymatically with lysinc residues, leading to Schiff-base

intermediates that react to stable Amadori products, the glycated

proteins (39). There also are experiments indicating that soluble

proteins within the peritoneal cavity can be modified by glycation

during CAPD treatment. This seems not to be the case for CIP,

because we could isolate CIP from the ultrafiltrate of HD patients

as well. Furthermore, it could be shown that glycatcd proteins

preferentially accumulate in the CAPD effluent (40). Experiments

to identify the type of modification leading to CIP arc ongoing in

our laboratory.

We conclude from the results presented here that a modified

form of ubiquitin (CIP) exists in the pcnitoneab cavity of PD

patients and that this modification is responsible for the inhib-

itory effect of CIP on PMNL cheruotaxis and, hence, for the

increased risk of infection, leading, at least in part, to a

higher incidence of morbidity and mortality in this group of

patients.

References1 . Mailloux LU, Bellucci AG, Wilkes BM, Napolitano B, Mossey RT,

Lesser M, Bluestone PA: Mortality in dialysis patients: Analysis of

the causes of death. Am J Kidney Dis 18: 326-335, 19912. Korbet SM, Vonesh EF, Firanek CA: A retrospective assessment

of risk factors for peritonitis among an urban CAPD population.

Peril Dial mt 13: 126-131, 1993

3. Vanholder R, Van Loo A, Dhondt AM, Dc Smet R, Ringoir 5:Influence of uraemia and haemodialysis on host defence and

infection. Nephrol Dial Transplant I I : 593-598, 1996

456 Journal of the American Society of Nephrology

4. Himmelfarb I, Hakim RI: Biocompatibility and risk of infectionin heruodialysis patients. Nephrol Dial Transplant 9[Suppl 2]:

138-144, 1994

5. Haag-Weber M. H#{246}rlWH: Dysfunction of polymorphonuclear

leukocytes in uremia. Semi?l Nephrol 16: 192-201, 19966. Ho-Dac-Pannekeet MM, Krediet RT: Inflammatory changes in

vivo during CAPD: What can the effluent tell us. Kidney hit

50[Suppl 56]: 512-516, 1996

7. H#{246}rlWH, Haag-Weber M, Georgopoulos A, Block LH: Physi-

cochemical characterization of a polypeptide present in uremic

serum that inhibits the biological activity of polymorphonuclear

cells. Proc Natl Acad Sci USA 87: 6353-6357, 1990

8. Cohen G, Haag-Weber M, Mai B, Deicher R, H#{246}rlWH: Effect of

immunogbobulin light chains from hemodialysis and continuous

ambulatory peritoneal dialysis patients on polymorphonuclearleukocyte functions. J A,n Soc Nephrol 6: 1592-1599, 1995

9. Haag-Weber M, Mai B, H#{246}rlWH: Isolation of a granubocyte

inhibitory protein in uremia: Homology to beta,-microgbobulin.

Nephrol Dial Transplant 9: 382-388, 1994

10. Haag-Weber M, Mai B, H#{246}rlWH: Impaired cellular host defencein peritoneal dialysis by two granulocyte inhibitory proteins.

Nephrol Dial Transplant 9: 1769-1773, 1994

1 1 . Tschcsche H, Kopp C, H#{246}rlWH, Hempelmann U: Inhibition of

degranulation of polymorphonuclear leukocytes by angiogenin

and its tryptic fragments. J Biol Chein 269: 30274-30280, 1994

12. Balke N, Holtkamp U, H#{246}rlWH, Tschesche H: Inhibition of

degranulation of human polymorphonuclear leukocytes by com-

plement factor D. FEBS Lett 371: 300-302, 1995

13. Haag-Webcr M, Cohen G, H#{246}rlWH: Modulation of granubocyte

(PMNL) function by complement factor D isolated from perito-

neal dialysis effluent [Abstract]. JAm Soc Nephrol 7: 1480, 1996

14. Vanholder R, Dc Smet R, Waterloos MA, Van Landschoot N,

Vogeleere P, Hoste E, Ringoir 5: Mechanisms of uremic inhibi-

tion of phagocyte reactive species production: Characterizationof the role of p-cresol. Kidney mt 47: 510-517. 1995

15. Metcalf IA, Gallin II, Nauseef WM, Root RK: Laboratory Man-

ual of Neutrophil Function, New York, Raven, 1986, pp 2-5

16. Nelson RD. Quie PG. Simmons RL: Chemotaxis and spontane-

ous migration of human polymorphonuclear leukocytes and

monocytes. J Immunol 1 15: 1650-1655, 1975

17. Schlesinger DH, Goldstein G, Niall H: The complete amino acid

sequence of ubiquitin, an adenylate cyclase stimulating polypep-

tide probably universal in living cells. Biochemistry 14: 22 14-

2218, 1975

18. Vijay-Kumar 5, Bugg CE, Cook WI: Structure of ubiquitinrefined at 1.8 A resolution. J Mol Biol 194: 5 13-544, 1987

19. Hans AL, Murphy KE, Bright PM: The inactivation of ubiquitin

accounts for the inability to demonstrate AlP, ubiquitin-dependent

proteolysis in liver extracts. J Biol C’hem 260: 4694-4703, 1985

20. Vanholder R, Van Loo A, Dhont AM, Dc Smet R, Ringoir 5:

Influence of uraemia and haemodialysis on host defence andinfection. Nephrol Dial Transplant 1 1 : 593-598. 1996

21 . Vanholder R, Lameire N, Waterloos MA, Van-Landschoot N,

De-Smet R. Vogeleere P. Lambert MC, Vijt D, Ringoir 5:

Disturbed host defence in peritoneal cavity during CAPD: Char-

acterization of responsible factors in dwell fluid. Kidney mt 50:

643-652, 199622. Porter Cl, Burden RP, Morgan AG, Daniels I, Fletcher I: Im-

paired polymorphonuclear neutrophil function in end-stage renal

failure and its correction by continuous ambulatory peritoneal

dialysis. Nephron 71: 133-137, 1995

23. Leypoldt 1K, Cheung AK: Removal of high-molecular-weightsolutes during high-efficiency and high-flux haemodialysis.

Nephrol Dial Transplant 1 1: 329-335, 1996

24. Downey GP: Mechanisms of leukocyte motility and chemotaxis.

Curr Opin bminunol 6: 1 13-124, 199425. Baggiolini M, Walz A, Kunkel SL: Neutrophil-activating peptide

1/interleukin 8, a novel cytokine that activates neutrophils. J Cliii

invest 84: 1045-1049, 1989

26. Marasco WA, Phan SH, Krutzsch H, Showell Hi, Feltner DE,

Nairn R, Becker EL, Ward PA: Purification and identification of

formyl-methionyl-leucyl-phenylalanine as the major peptide

neutrophil chemotactic factor produced by E.s-cherichia coli.

J Biol Chem 259: 5430-5439, 1984

27. Goldstein IM: Cheruotactic factor receptors on leukocytes:

Scratching the surface. J Lab Clin Med 97: 599-601, 1981

28. Bos HI, Boorsma DM, Tuk CW. Dc Veld IC, Van Der Muysen-

berg AJC, Helmcrhorst TIM, Struijk DG, Van Bronswijk, Beelen

RHI: Chemotaxis of the penitoneal cells and the detection of a

chemo-attractant in the effluent from penitoneal dialysis patients.

Eur J Cliii bnvest 20: 555-562, 1990

29. Schlesinger DH, Goldstein G: Molecular conservation of 74

amino acid sequence of ubiquitin between cattle and man. Nature

255: 423-424, 1975

30. Goldstein G, Scheid M, Hammerling U, Boyse EA, Schlesinger

DH, Niall HD: Isolation of a polypeptide that has lymphocyte-

differentiating properties and is probably represented universally

in living cells. Proc NatI Acad Sci USA 72: 1 1-15, 1975

3 1 . Hershko A: Ubiquitin-mediated protein degradation. J Biol Chem

263: 15237-15240, 1988

32. Chau V. Tobias 1W, Bachruair A, Marriott D, Ecker Di, Gonda

DK, Varshavsky A: A multiubiquitin chain is confined to spe-

cific lysine in a targeted short-lived protein. Science 243: 1576-

1583, 1989

33. Mitch WE, lurkovitz C, England BK: Mechanisms that cause

protein and amino acid catabolism in uremia. Am J Kidney Dis

21: 91-95, 1993

34. Mitch WE, Goldberg AL: Mechanisms of muscle wasting: The

role of the ubiquitin-proteasorue pathway. N Engl J Med 335:

1897-1905, 1996

35. Pancr#{233}V. Pierce RI, Foutnier F, Methali M, Delanoye A, Capron

A, Auriault C: Effect of ubiquitin on platelet functions: Possibleidentity with platelet activity suppressive lymphokine (PASL).

Ear J Immunol 2 1 : 2735-274 1 , 1991

36. Okada M, Miyazaki 5, Hirasawa Y: Increase in plasma concentra-

tion of ubiquitin in dialysis patients: Possible involvement in beta2-

microglobulin amyboidosis. Cliii Chi,n Ada 220: 135-144, 1993

37. Roxborough HE, Millar CA, McEneny I, Young IS: Carbamy-

lation inhibits the ferroxidase activity of caeruboplasmin. Bio-

chein Biophys Res Coinmun 2 14: 1073-1 078, 1995

38. Stark GR, Stein WH, Moore 5: Reactions of the cyanate present

in aqueous urea with amino acids and proteins. J Biol Chem 235:

3177-3181, 1960

39. Schliecher E, Wieland OH: Protein glycation: Measurement and

clinical relevance. J Cliii Chem Cliii Biochem 27: 577-587, 1989

40. Lamb E, Cattel WR, Dawnay A: Glycated albumin in serum and

dialysate of patients on continuous ambulatory penitoneal dialy-

sis. Cliii Sci 84: 619-626, 1993