Download The Urinary F1 Activation Peptide of Human

533
Clinical Science (1995) 89, 533-541 (Printed in Great Britain)
The urinary FI activation peptide of human prothrombin
is a potent inhibitor of calcium oxalate crystallization in
undiluted human urine in vitro
Rosemary L. RYALL, Phulwinder K. GROVER. Alan M. F. STAPLETON, Dianne K. BARRELL,
Yulu TANG. Robert L. MORITZ* and Richard J. SIMPSON*
Department of Surgery, Flinders Medical Centre, Bedford Park, South Australia, Australia, and
*Joint Protein Structure Laboratory, Ludwig Institute for Cancer Research (Melbourne Branch)
and the Walter and Eliza Hall Institute of Medical Research, Parkville, Victoria, Australia
(Received 18 April/28 June 1995; accepted 12 July 1995)
1. The urinary F1 activation peptide of prothrombin
is the predominant protein incorporated into calcium
oxalate crystals precipitated from human urine. The
aim of this study was to examine the effect of pure
urinary prothrombin F1 on calcium oxalate crystallization in human urine.
2. Urinary prothrombin F1 was purified from demineralized calcium oxalate crystals precipitated from
human urine, and its effects on calcium oxalate
crystallization induced by addition of an oxalate load
were tested in undiluted, ultrafiltered urine from
healthy men, at final concentrations of 0 to 10 mgjl.
3. Urinary prothrombin F1 did not affect the amount
of oxalate required to induce crystallization, but the
volume of material deposited increased in proportion
to increasing concentrations of urinary prothrombin
Fl. However, the mean particle size decreased in
reverse order: this was confirmed by scanning electron
microscopy, which showed it to be the result of a
reduction in crystal aggregation rather than in the
size of individual crystals. Analysis of 14C-oxalate
data revealed a dose-dependent decrease in calcium
oxalate deposition with an increase in urinary prothrombin F1 concentration, indicating that the
increase in particle volume recorded by the Coulter
Counter resulted from inclusion of urinary prothrombin F1 into the crystalline architecture, rather than
increased deposition of calcium oxalate.
4. It was concluded that urinary prothrombin F1 may
be an important macromolecular determinant of
stone formation.
INTRODUCTION
A possible role for urinary proteins in the genesis
of kidney stones was proposed more tha~ 30 yea~s
ago when it was shown that the orga~llc matnx
known to be present in all renal calcuh [1] con-
sisted principally of protein [2]. However, in the
absence of any conclusive evidence to confirm or
refute the existence of such a role, urinary proteins
were largely ignored in this respect, until more
recently, when the advent of more sophisticated
techniques for isolating, purifying, analysing and
characterizing proteins, rekindled interest in the
possibility that they may influence the course of
stone disease. To avoid the well-recognized difficulties associated with studying the proteins of stone
matrix [3], several investigators have focused on
those present in calcium oxalate (CaOx) crystals
newly precipitated from fresh human urine [3-5].
These studies indicated that, despite the hundreds of
proteins known to be present in human urine,
relatively few actually participate directly in the
nucleation and growth of CaOx crystals: the organic
matrix of such crystals contains only specific urinary
proteins. The selectivity of the inclusion of the~e
proteins into the organic matrix of such crystals .IS
demonstrated by the fact that they are present m
quantities apparently unrelated to their abunda~~e
in the urine from which the crystals were precipitated. For instance, the most plentiful protein in
human urine, Tamm-Horsfall glycoprotein (THG),
if present in the matrix at all, is detectable in only
very minor amounts [3]. On the other hand, one
particular protein is present in quantities which
exceed those of all others, and which belie its low
concentration in urine. Originally named crystal
matrix protein, because early studies were unable to
demonstrate its identity with known urinary proteins [3], it is now known as urinary prothrombin
Fl (UPTFl), after the demonstration of its relationship to human prothrombin [6, 7] and, more
recently, confirmation of its close identity .with the
Fl activation peptide of human prothrombin [8, 9].
The fact that UPTFI is the principal component
of CaOx crystal matrix extract (CME), which is a
potent inhibitor of CaOx crystal growth and aggre-
Key words: prothrombin activation peptide FI, calcium oxalate, urinary inhibitors, crystal growth and aggregation, urolithiasis..
. . . UPTFI
. r
Abbreviations: CaOx, calcium oxalate; CME, crystal matrix extract; HSA, human serum albumin; THG, Tamm-Horsfall glycoprotein; UF, ultrafiltration,
, unna y
prothrombin FI.
.
. '
Correspondence: Dr R. L. Ryall, Department ofSurgery, Hinders Medical Centre, Bedford Park, South Australia 5042, Australia.
534
R. L. Ryall et al.
gation in urine [10], its presence in human kidney
stones [8, 9] and its precise distribution within the
human nephron [II], suggests that under normal
physiological conditions it may fulfil a determinant
function as an inhibitor of CaOx crystallization, and
thereby of stone formation. However, to date, it has
not been possible to obtain experimental evidence
to support this hypothesis, because insufficient protein has been available to determine whether it can
account for the inhibitory potency of the organic
matrix of CaOx crystals. The aim of this study was
to isolate and purify UPTFI from human urine in
quantities large enough to allow the measurement of
its inhibitory effect on CaOx crystallization in undiluted urine in vitro.
MATERIALS AND METHODS
Preparation of CME
Collection and preparation of urine samples. Urine
samples (24 h) were collected from healthy men who
had no history of kidney stone disease. The samples
were refrigerated during the collection period and
during storage before use. Absence of blood from
the specimens was confirmed using Multistix test
strips (Miles Laboratories, Mulgrave, Victoria, Australia). The samples were pooled and centrifuged at
80001: for 15 min at 20°C in a Beckman 12-21 M/E
centrifuge (Beckman Instruments, Palo Alto, CA,
U.S.A.). The supernatant was filtered through
O.22flm Millipore filters (# GVWP 14250, Millipore Corporation, Bedford, MA, U.S.A.).
Generation of CaOx crystals. Ca Ox crystals were
precipitated from the pooled urine as previously
described [3]. Briefly, the metastable limit of the
urine with respect to CaOx, i.e. the minimum
amount of oxalate required to elicit spontaneous
detectable CaOx crystallization using the Coulter
Counter, was first determined by titration with
sodium oxalate solution. A standard load of oxalate
(30flmo1/100ml urine) in excess of the metastable
limit was then added with continuous stirring to the
sample in 51 volumetric glass flasks, and the flasks
were shaken at 900pm in a water bath heated to
37°C. To increase the yield of crystals, the same
amount of sodium oxalate solution was added after
I h and again after 2 h. The total incubation time
was 4h.
The crystals were harvested by filtration through
0.22 flm Millipore filters and washed thoroughly
with 0.1 mol/l sodium hydroxide followed by distilled water (Permutit Australia, Brookvale, NSW,
Australia). They were then lyophilized and weighed.
Purification of UPTFI
To every I g of CaOx crystals was added 100ml
of 0.25 mol/I EDTA solution (pH 8.0), and the
suspension was stirred at room temperature until
crystals were no longer detectable by light micro-
scopy. The resulting CME, prepared from several
batches of urine, was pooled, lyophilized and dissolved in 25 mmol/l Tris-HCI butTer (pH 7.3). It was
then applied to a Bio-Gel P-6 DG desalting column,
which was developed with the same butTer. The
protein peak was collected manually and a small
sample was subjected to an analytical run on
reversed-phase HPLC using a Brownlee RP-300
microbore column (2.1 mm x 100 mm). A 30 min
linear gradient was developed from 0 to 100% B,
where solvent A was 0.1% aqueous trifluoroacetic
acid and solvent B was 0.09% trifluoroacetic acid in
60% acetonitrile, 40% water. The column temperature was ambient (approximately 21 DC) and the flow
rate was 0.2 ml/min, The eluant was monitored by
absorbance at 215 nm. The protein peak corresponding to what was presumed to be UPTFI on
the basis of its relative retention behaviour [6] was
collected and analysed by SDS/PAGE, which confirmed that the sample consisted almost exclusively
of a single band of UPTFI, as judged by its
molecular mass of 31 kDa. A larger, preparative
reversed-phase HPLC run was then performed with
the remainder of the sample from the desalting
column, using identical conditions to those described for the analytical run, except that a Vydac
C4 preparative column (lOmm x lOOmm) was used
with a flow rate of 10 ml/min. Fractions were
collected at I min intervals. The fractions known to
contain UPTF1 were collected and dialysed extensively against phosphate butTer (pH 7.3), containing
NaCi 8.5 gil, K 2HP04 1.43 gil and KH 2P04
0.25 gil, at 4°C and the protein concentration was
determined using the fluorescamine method described by Bohlen et al. [12] and modified by Castell
et al. [13]. The proteins were hydrolysed by autoc1aving in I mol/l NaOH at 120°C for 70 min before
the assay. Both BSA and a-lactalbumin were used
as standards.
50S/PAGE
The samples to be analysed were reduced with
2-mercaptoethanol and electrophoresed in 1 mmthick, 10-20% linear gradient gels on a Bio-Rad
Mini-Protean II apparatus as previously described
[3], using the method based on principles established by Laemmli [14]. The gels were subsequently
stained with silver [15].
Determination of inhibitory activity
Urine samples (24 h) were collected from five
healthy men, and were pooled, centrifuged and
filtered as described above. Sodium azide was added
at a final concentration of 0.025% and a portion
was retained as a control (centrifuged and filtered
sample). This control contained its full complement
of macromolecules other than THG, which is
known to be removed from urine by centrifugation
and filtration [16, 17]. The remaining urine was
Urinary FI activation peptide and calcium oxalate urolithiasis
535
ultrafiltered (nominal molecular mass cut-off of
10kOa) using an Amicon hollow fibre bundle
(HI PIO-20; Amicon Corporation, Lexington, MA,
U.S.A.) and divided into aliquots, to which purified
UPTFI was added at final concentrations of 0, 1.0,
2.5 and 10mg/l. Inhibitory activity was assessed as
previously described [18]. The metastable limit of
all the portions was determined by titrating 4 ml
samples with sodium oxalate solution as described
above, using the Coulter Counter to detect crystals
> 2 J-lm. Once the metastable limit had been determined, a standard load of oxalate (30 J-lmoljl00ml)
in excess of this amount was added dropwise to the
samples which were then incubated in a shaking
water bath at 37°C for 120min. Crystal size distributions were measured at 30 min intervals using the
Coulter Counter, and from these were calculated the
total volumes of particulate material precipitated
during the incubation period and the modes of the
volume-size distribution curves. Because quantities
of purified UPTFI were limited, each experiment
could only be performed in duplicate; results are
expressed as the mean of the two measured values.
Because preliminary studies had confirmed that
UPTFI did not precipitate during the course of the
experiment, control flasks containing all components other than oxalate were not included to
account for possible counting interference caused by
precipitated protein.
mass, parallel incubations were carried out with
samples containing 14C-oxalate. In these samples,
any decrease in soluble radioactivity must reflect
precipitation of CaOx.
The radioactive samples were treated identically
to the ones described above, except that 14C-oxalic
acid (l0 J-lCij100ml) was added before induction of
CaOx crystallization by addition of the oxalate
load. At 30 min intervals, 1.0ml samples were filtered (0.22 J-lm) into 100J-li of concentrated HCl
using a disposable syringe and filter holder. Filtration into acid was a precaution against the possibility that further precipitation of CaOx might
occur after filtration of the solution. Duplicate
0.3 ml portions of the acidic solution were then
added to 10ml of Ready Safe (Beckman Instruments) and counted for 5 min in a liquidscintillation counter (Beckman LS 3801 Liquid Scintillation System). Once again, each experiment was
performed in duplicate and the results are expressed
as mean values.
Scanning electron microscopy
RESULTS
At the end of each experiment, 2 ml aliquots were
filtered (0.22 J-lm filters) and the filters were dried
overnight at 37°C, after which they were mounted
on aluminium stubs and gold-spluttered for 180s
(SEM Autocoating Unit E5200, Polaron Equipment
Ltd, Watford, U.K.). Stubs were examined using an
ETEC Auto Scan Electron Microscope (Siemens
AG, Karlsruhe, Germany) at an operating voltage
of 20 kV, with the operator blinded with regard to
the experimental source of the crystals.
CME and purification of crystal matrix protein
Measurement of CaOx deposition by '4C-oxalate
There are recognized limitations associated with
using the Coulter Counter to determine the volume
of particles, especially if those particles consist of
crystal aggregates: the Coulter Counter will record
the interstices between individual crystals as solid
crystalline material. Any error this may cause will
be compounded by the fact that the total volume of
crystals precipitated from urine comprises solid
CaOx as well as urinary macromolecules, including
UPTFl, trapped inside the crystalline architecture
[10]. Furthermore, the Coulter Counter measures
particles whose sizes fall within specified limits;
crystals falling outside this size range (2-25.4 J-lm in
these experiments) will not be counted by the
instrument. Therefore, to determine whether particle
volume accurately reflected the increase in crystal
Ethical considerations
This study involved collection of urine samples
from healthy control subjects, and was reviewed and
approved by the Committee on Clinical Investigation (Ethics Committee) of the Flinders Medical
Centre (minute no. 3024.12, 1993).
Figure 1 shows the SOS/PAGE electrophoretogram of the pooled desalted CME. Note that
although the predominant protein band is UPTFI,
with an apparent molecular mass of 30 kOa, other
bands are also visible, including urinary human
serum albumin (HSA) with an apparent molecular
mass of 67 kOa and other, as yet unidentified,
proteins. Figure 2 shows the elution profile of the
extract after reversed-phase HPLC, which reveals
the presence of two major peaks, the larger of which
was shown to be urinary albumin by immunoblotting with an anti-HSA antibody, as well as a
number of minor peaks, confirming the SOS/PAGE
findings. Figure 3 shows the SOS/PAGE analysis of
the peak corresponding to UPTFl, demonstrating
that it contained almost pure UPTF1. It was therefore used in the inhibition experiments without
further purification.
Effect of UPTFI on the urinary metastable limit
with respect to CaOx. The minimum amount of
oxalate required to elicit spontaneous detectable
CaOx crystal nucleation was unaltered by ultrafiltration and by addition of increasing amounts of
UPTFI up to and including a final concentration
of IOmg/1. Therefore, an identical oxalate load
was added to all the samples to induce CaOx
crystallization.
R. L. Ryall et al.
536
Fig. 3. 50S/PAGE gel (1G-20%) of peak (*) from reversed-phase
HPLC. demonstrating the presence of almost pure UPTFI
Fig. I. 50S/PAGE gel (9-18%) of proteins obtained by demineralization of CaOx crystals precipitated from fresh undiluted urine
collected from healthy men. Lane 2 represents standard molecular
mass markers
25000
=~
.~
-~
1.0
ro
0 ..····
~
0.75
~
--0--
10000
----6----
UF + 1.25 mg(1 UPTFI
UF + 2.5 mgjl UPTFI
UF + 10 mgjl UPTFI
~
E
~
~ 0.50
j
-{}-- UF
..30 15000
>0-
I
20000
E
~
- - -ffi- - -
5000
Centrifuged and
filtered urine
0.25 -
~
-c
30
- - - - -....
_---
Fig. 2. Reversed-phase HPLC chromatogram of the crystal matrix
proteins shown in Fig. I. The asterisk indicates the peak containing
UPTFI.
60
90
120
Time (min)
Fig. 4. The increase in deposition of particle volumewith time after
the addition of the oxalate load in centrifuged and filtered urine,
and in the same urine after ultrafiltration (UF) at a nominal
molecular mass cut-off of 10kOa. to which UPTFI had been added
at final concentrations of 0, 1.0,2.5 and 10mg/1
Effect of UPTF I on particle volume
Figure 4 shows the time course of total particle
volumes precipitated during the 120min incubation
period after addition of oxalate load to the centrifuged and filtered control and the ultrafiltered urine
samples containing varying amounts of UPTFI. In
all cases, there was an initial lag phase, followed by
a period of near-linear volume deposition which
then approached a plateau. It is noteworthy that
from zero time, the results for the centrifuged and
filtered urine differed markedly from those obtained
for all the ultrafiltered urine samples, which were
superimposable for the first 30 min and then barely
distinguishable from each other up to 60 min. After
that time, values obtained in the presence of UPTFl
were clearly different from those obtained in the
ultrafiltered urine alone, although differing only
slightly from each other. The mean volume of
particulate material deposited after 120 min in the
centrifuged and filtered urine was 23450.um 3/.ul,
while the corresponding values in the ultrafiltered
urine samples were 8154, 14795, 16096, and
16814.um3/.u1 at final UPTFI concentrations of 0,
1.25, 2.5 and 10 rng/l respectively. These represent
respective particle volume increases of 81.44, 97.47
and 106.20(~~. in proportion to the ultrafiltered
control containing no UPTF 1.
Urinary FI activation peptide and calcium oxalate urolithiasis
--0-
UF
······-0·····...
UF + 1.25 mg/l UPTF I
···-0-···
UF+2.5mg/l UPTFI
-·--6---- UF+ IOmg{1 UPTFI
---m- .... Centrifuged and
filtered urine
Fig. 5.The particlevolume distribution at 120min after the addition
of the oxalate load in centrifuged and filtered urine, and in aliquots
of the sameurine which had beenultrafiltered(UF) throughhollow
fibres with a nominal molecular mass cut-off of 10 kDa and spiked
with 0, 1.0,15 and 10mgli of UPTFI
Effect of UPTFI on particle size
Figure 5 shows the size of the particles precipitated from the various urine samples, expressed as
the modes of the volume distribution curves at the
end of the 120min incubation period. The most
notable feature of these results is that the position
of the modes of the curves varied directly in response to the concentration of UPTFI in the ultrafiltered urine. Thus, the mean particle size in the
ultrafiltered urine containing no UPTFI was
18.3 j.lm, and this decreased steadily to 14.2, 12.7
and 7.3 j.lm at corresponding UPTFI concentrations
of 1.25, 2.5 and 10 mg/l. It is remarkable that the
size of the particles deposited from the ultrafiltered
urine containing the highest concentration of
UPTF1 (7.3 j.lm) was less than that of those precipitated from the centrifuged and filtered control urine
(8.5 j.lm), which contained a normal complement of
urinary macromolecules (including UPTF1) other
than THG. These findings were confirmed by scanning electron microscopy.
Figure 6 shows low-power scanning electron micrographs of crystals deposited from the centrifuged
and filtered control and the ultrafiltered urine samples containing varying concentrations of UPTFl.
Mainly CaOx dihydrate crystals were precipitated:
those from the centrifuged and filtered urine tended
to be single or clustered into small loose aggregates.
This sample also showed the presence of precipitated organic material on the filter surface. In
contrast, the crystals precipitated from the ultrafiltered urine containing no added UPTF1 were
highly aggregated and no precipitated organic
material was observed. The degree of aggregation of
the precipitated crystals decreased in relation to
increasing concentrations of UPTF1, indicating that
the dose-dependent reduction in particle size occurring with increasing concentrations of UPTFI noted
above, was the result of a decrease in the degree of
crystal aggregation, rather than in the size of individual crystals. In fact, it was apparent from obser-
537
vation of a large number of crystal fields that the
presence of UPTFI tended to be associated with
some very large individual crystals, the number of
which increased in proportion to the concentration
of the protein. This is illustrated in the micrographs
presented. Furthermore, the surfaces of most of the
crystals precipitated in the presence of UPTF1, but
particularly the larger ones, have a mottled or
mosaic appearance: the effect is more obvious in
samples containing high concentrations of UPTFl,
especially at high magnification, as shown in Fig. 7.
This electron micrograph shows examples of the
larger crystals precipitated from the plain ultrafiltered urine, that to which 10mg/1 UPTF1 had been
added, and the same urine which had been simply
centrifuged and filtered. Neither the crystals precipitated from the plain ultrafiltered urine nor those
from the centrifuged and filtered urine exhibited the
mosaic patterns seen on those from the ultrafiltered
urine containing UPTFl. This suggests that,
although the effect is probably the result of binding
of macromolecules to the surfaces of the crystals, it
can be attributed specifically to the binding of
UPTF1, otherwise the pattern would have been
observed on the crystals from the centrifuged and
filtered urine.
Measurement of CaOx deposition by 14C-oxalate
Figure 8 shows the time course of the disappearance of 14C-oxalate from the supernatants of the
urine samples during the 120min incubation period
after addition of the oxalate load. To normalize the
data, the values are presented as the amount of
14C-oxalate remaining in solution, expressed as a
percentage of the zero time value. There are two
remarkable features. Firstly, the rate of disappearance of 14C-oxalate from the centrifuged and filtered
urine was greater than in any of the ultrafiltered
samples. Secondly, among the various ultrafiltered
urine samples, the rate of reduction in 14C-oxalate
decreased in proportion to increasing concentrations
of UPTF1. It should be noted that the data demonstrate that the deposition of 14C-oxalate in the
centrifuged and filtered urine was greater than in
any of the ultrafiltered urine samples: at 120min,
the mean amount of 14C-oxalate remaining in the
control urine (51.1%) was less than that in the
ultrafiltered urine containing no added UPTFI
(59.4%). The corresponding values in the ultrafiltered urine samples containing UPTFI at final
concentrations of 1.25, 2.5 and 10 mg/l were 74.0%,
78.4% and 77.8% respectively, corresponding to
degrees of inhibition of CaOx deposition of 24.7%,
32.2% and 31.1% compared with ultrafiltered urine
alone. This demonstrates that the dose-dependent
increase in particle volume that was recorded by the
Coulter Counter with increasing concentrations of
UPTFI could not be attributed to increases in the
deposition of CaOx. On the contrary, the 14C_
oxalate data indicate quite clearly that CaOx
538
R. L. Ryall et al.
Fig. 6. Low-power scanning electron micrographs of crystals precipitated in urine that had been centrifuged and filtered
(labelled here as spun and filtered), and in portions of the same urine which had been ultrafiltered (UF) through hollow
fibres with a nominal molecular mass cut-off of 10kDa and supplemented with 0, 1.0,2.5 and 10mg/i of UPTFI
deposition is inhibited, not promoted, in proportion
to the prevailing concentration of U PTF 1.
DISCUSSION
Previous studies have shown that UPTFI possesses all the characteristics expected of a urinary
macromolecule fulfilling a directive function in
CaOx stone pathogenesis. The protein is present in
calcium stones [9], is confined to specific, circumscribed regions of the human nephron [II] and is
the principal protein present in the organic matrix
of CaOx crystals precipitated from fresh human
urine [3]. This organic matrix, CME, has been
shown to be the most potent inhibitor of CaOx
crystallization induced in human urine so far described [10]. However, although these features, collectively, suggest that UPTFI might be involved in
controlling the crystallization of CaOx ill rioo, the
evidence for its fulfilling such a function has been, at
best, indirect, since the protein has not been available in sufficient quantities free of other macromolecular contaminants to test its inhibitory activity in
urine. It was the object of this study to determine
whether UPTFI purified from human urine could
account for the observed potent inhibitory effect of
CaOx crystal matrix on the nucleation, growth and
aggregation of CaOx crystals in undiluted urine,
and thereby, to establish whether the protein fulfils
some function in stone formation.
UPTFI was isolated from CaOx crystals precipitated from male human urine in large-scale crystallization experiments using methods already established in our laboratory, and then purified from the
pooled, demineralized extract of these crystals by
reversed-phase HPLC to yield a homogeneous preparation that gave a single band on SDS/PAGE.
This material was considered sufficiently pure to be
used in the inhibition studies. The results of these
experiments were unequivocal; UPTFI is a potent
inhibitor of CaOx crystal growth and aggregation in
undiluted urine, but like all macromolecules previously tested with the technique used here [10, 1921], did not affect the amount of oxalate required to
induce detectable spontaneous nucleation of CaOx.
Furthermore, as was shown previously with CME
[10], the volume of precipitated material increased
in a dose-response fashion in the presence of the
Urinary FI activation peptide and calcium oxalate urolithiasis
539
Fig. 7. High-power scanning electron micrographs of the same crystals shown in Fig. 6, derived from centrifuged and
filtered urine and ultrafiltered (UF) urine containing 0 and 10mg!1 of UPTFI
100 -
§:
.
Q)
90
:;;
..
,,~~,
~~::'·:::~~:~·'o -,
<~::...
x
'i' 80
~
1l
~.
·0······
"0
-··-0····
~ 70
'Q.
.~
g- 60
::::>
-o-UF
>,'9
"<,
----6----
.lB..••••
-- -1Il-- -
UF + 1.25 mg/I UPTFI
UF + 2.5 mgll UPTFI
UF + 10mg/1 UPTFI
Centrifuged and
filtered urine
"&1.,
50 ........,.--.,.--.,......-..,..--11130
60
90
120
Time (min)
Fig. 8. Percentage decrease in soluble "C-oxalate with time after
the addition of the oxalate load in centrifuged and filtered urine,
and in portions of the same urine which had been ultrafiltered at a
nominal molecular masscut-off of 10kDa and supplemented with 0,
1.0, 2.5 and 10mg/i of UPTFI
protein, despite the fact that the amount of
14C-oxalate decreased in exact reverse order. A
similar, apparently contradictory observation has
been reported before: CME significantly increased
the volume of precipitated material, while having no
effect on the deposition of 14C-oxalate [10]. In that
instance, the observed increase in particle volume in
the absence of a corresponding promotion of
14C-oxalate deposition, was attributed to the inclusion of urinary macromolecules into the crystalline
architecture. It is apparent that although included
organic material accounts for only a small proportion of a crystal's (and indeed a stone's) mass,
nonetheless, it can occupy a significant volume of
the structure. It is unlikely that the apparent discrepancy between the Coulter Counter and 14C-oxalate
data observed in the present study can be attributed
simply to shortcomings associated with the technique of particle size analysis, although this may
have been partly responsible. The Coulter Counter
estimates a particle's size on the basis of the volume
of electrolyte it displaces as it traverses the counting
orifice; it cannot distinguish between particles of
different composition or density. Thus, a loosely
aggregated cluster of crystals, though containing
large lacunae filled with the urine in which it is
suspended, will be recognized by the Coulter
Counter simply as a large solid particle, thereby
overestimating the volume of crystalline material in
suspension. If this were the sole explanation for the
apparent incompatibility between the volume and
14C-oxalate data, then we would have expected the
measured volume to decrease as the degree of
aggregation of the crystals reduced in relation to the
concentration of UPTFl, not increase, as was
observed. Thus, although drawbacks accompanying
the Coulter Counter assessment of particle volume
may partly account for the apparently inconsistent
findings, a more probable explanation is that macromolecules trapped within the crystals occupy a
significant proportion of the structure's volume and
thereby increase its overall size. Certainly, this
would explain the empirical observation of an
increased number of large individual crystals precipitated from the urine samples containing UPTFI.
Nonetheless, while UPTFI increased the overall
bulk of particulate material precipitated, it effectively reduced the amount of CaOx deposited.
UPTFI concentrations of 2.5 and lOmgjl inhibited
the deposition of 14C-oxalate by approximately
30%. The protein is therefore an efficient inhibitor of
crystal growth, where this is defined as the increase
R. L. Ryall et al.
in size brought about specifically by the deposition
upon the crystal surface of new solute ions, and in
this respect it differs from other macromolecules
whose effects on CaOx crystal growth in urine have
been similarly measured. Although numerous studies have reported that a number of urinary proteins inhibit the deposition of CaOx from inorganic
solutions [21], to our knowledge, the only individual urinary proteins whose effects on CaOx crystallization have been tested in undiluted urine are
HSA [21] and THG [21-23], and of these, the
effects of only THG have been measured using a
combination of both Coulter Counter and
14C-oxalate analysis [23]. At a concentration of
20mg/l, HSA slightly, but significantly, increased the
volume of CaOx precipitated [21]. This observation
is in keeping with the fact that the protein is
incorporated into CaOx crystals precipitated from
urine, although in quantities less than might be
expected, in view of its abundance in urine [3]. On
the other hand, THG quite clearly promoted the
deposition of CaOx at unusually high urine osmolalities [22, 23], but had no effect at more common
physiological osmolalities [23]. Thus, to date, no
naturally occurring individual urinary protein, with
the notable exception of UPTFl, nor indeed any
mixture of proteins [10], has been demonstrated to
inhibit significantly the deposition of CaOx in undiluted urine. However, although UPTFI is the first
urinary protein with demonstrable inhibitory effects
on CaOx crystal growth, it is its effect on crystal
aggregation that is likely to be of greater physiological relevance in the prevention of urolithiasis.
Crystal aggregation is well recognized to be a
crucial step in the formation of calculi, since it is the
only process by which large particles can be formed
in a short space of time. Over a concentration range
of 1.25 to l Omg/I, UPTFI inhibited the aggregation
of CaOx crystals that occurs upon removal of all
macromolecules with molecular masses greater than
IOkDa [19-21, 23]. The average size of the precipitated particles was reduced progressively from
18.3 JIm in the ultrafiltered urine to 14.2, 12.7 and
7.3 JIm at UPTFI concentrations of 1.25, 2.5 and
10 mg/l respectively. Most noteworthy, was the fact
that the particle size was reduced to less than that
of the particles precipitated from the centrifuged
and filtered urine, indicating that the endogenous
concentration of UPTFI in the particular centrifuged and filtered urine used in this study was less
than 10 mg/l, as would be expected from published
data. Using immunological techniques, Bezeaud and
Guillin [24] reported that the normal concentration
of Fl in urine ranged from 0.9 to 35.3 nmol/l, If we
assume a molecular mass of F 1 of approximately
30 kDa, these figures convert to approximately 0.03
to 1.1 mg/l, which is very close to the value of
1.25 mg/l used in the present study, at which inhibitory effects on both crystal aggregation and
14C-oxalate deposition were clearly evident. UPTFI
is therefore a powerful inhibitor of CaOx crystal
aggregation in undiluted urine at concentrations of
the order of those expected in normal urine, and
may thus play an important role in the prevention
of CaOx stone formation. Furthermore, the results
of this study demonstrate that UPTFI probably
accounts for the potent inhibition of crystal aggregation demonstrated by the organic matrix (CME)
of CaOx crystals precipitated from urine [10]. Only
one other protein, THG, has been shown to inhibit
CaOx crystal aggregation in undiluted ultrafiltered
urine [21, 23], but even at a final concentration of
20 mg/l, it was unable to reverse the formation of
large crystal aggregates that occurred upon ultrafiltration. Moreover, THG cannot account for the fact
that large crystal aggregates do not form in centrifuged and filtered urine, because it is efficiently
removed from urine by centrifugation and filtration
[16, 17]. Furthermore, it is virtually absent from
CME [3].
The strong inhibitory effects of UPTFI on CaOx
crystallization can probably be attributed to the
presence of its Gla domain which is the region
responsible for binding of the parent prothrombin
molecule to phospholipid membranes during blood
coagulation [25]. Although in the present study, the
strong binding of UPTFI to CaOx was indirectly
manifested as the reduction in 14C-oxalate deposition that occurred in the presence of the peptide,
binding was clearly visible from the scanning electron micrographs of the precipitated crystals. These
showed the presence of gross defects which
appeared as irregular mottled variegations covering
the crystal surfaces. Although visible on the smaller
crystals, the effect was most marked and conspicuous on the surfaces of the larger crystals. By
comparison, the margins of the crystals precipitated
from the ultrafiltered urine containing no UPTFI
were crisply delineated and the crystal surfaces
appeared smooth and unmarked. Crystals deposited
from the centrifuged and filtered urine, although
lacking the clean uniformity of those from the
ultrafiltered urine, and showing some obvious surface irregularities, nonetheless, did not exhibit the
extreme mosaic appearance of those deposited in
the presence of purified UPTFl, despite the fact
that the urine from which they had been derived
would have contained endogenous levels of the
peptide. Thus, the results presented here strongly
indicate that the exceptional inhibitory effects of
UPTFI on CaOx crystal growth and aggregation,
and its inclusion into those crystals, result directly
from its exceptional binding affinity for their surfaces, the strength of which almost undoubtedly can
be attributed to the y-carboxyglutamic acid residues
of the protein's Gla domain. However, although it
might be tempting to suppose that UPTFI, by
virtue of its Gla domain, is the principal macromolecular inhibitor of CaOx crystal aggregation in
urine, other published data highlight that such a
conclusion cannot be drawn at the present time.
Other urinary proteins, including nephrocalcin [26],
Urinary FI activation peptide and calcium oxalate urolithiasis
uropontin [27] and uronic-acid-rich protein (a close
relative of inter-a-trypsin inhibitor) [28] have been
shown to inhibit CaOx crystal growth in inorganic
solutions. It is possible, therefore, that these proteins
may also exert strong inhibitory effects in vivo, but
this cannot be assumed at present because their
effects on CaOx crystallization in undiluted human
urine have never been tested.
Nonetheless, irrespective of any inhibitory properties of other urinary proteins that may ultimately be
disclosed, the results of our investigation demonstrate that the urinary form of human prothrombin
fragment 1, UPTF1, is a potent inhibitor of CaOx
crystal growth and aggregation under conditions
approximating those operating in vivo. This finding
supports evidence already linking stone formation
and blood coagulation [8, 9], and reinforces the
possibility that UPTFI may play a critical role in
the prevention of urolithiasis.
ACKNOWLEDGMENTS
This work was supported by grants 940283 from
the National Health and Medical Research Council
of Australia and G21/95 from the Australian Kidney
Foundation. A.M.F.S. was the grateful recipient of a
Merck Sharp and Dohme Urological Research
Fellowship.
REFERENCES
I. Boyce WH, Garvey FK. The amount and nature of the organic matrix in
urinary calculi: a review. J Urol 1956; 76: 213-27.
2. King JS, Boyce WHo Analysis of renal calculous matrix compared with some
other matrix minerals and with uromucoid. Arch Biochem Biophys 1959; 82:
45>-61.
3. Doyle IR, Ryall RL, Marshall YR. Inclusion of proteins into calcium oxalate
crystals precipitated from human urine: a highly selective phenomenon. e1in
Chem 1991; 37: 1589-94.
4. Morse RM, Resnick MI. A new approach to the study of urinary
macromolecules as a participant in calcium oxalate crystallization. J Urol 1988;
139: 869-73.
5. Morse RM, Resnick MI. A study of the incorporation of urinary
macromolecules into crystals of different composition. J Urol 1989; 141: 641-4.
6. Stapleton AMF, Simpson RJ, Ryall RL. Crystal matrix protein is related to
human prothrombin. Biochem Biophys Res Commun 1993; 195: 1199-203.
7. Suzuki K, Moriyama M, Nakajima C, et al. Isolation and partial
characterization ofcrystal matrix protein as a potent inhibitor of calcium
oxalate crystal aggregation: evidence of activation peptide of human
prothrombin. Urol Res 1994; n: 45-50.
541
8. Stapleton AMF, Ryall RL. Blood coagulation proteins and urolithiasis are
linked: crystal matrix protein is the FI activation peptide of human
prothrombin. Br J Urol 1995; 75: 712-19.
9. Stapleton AMF, Ryall RL. Crystal matrix protein: getting blood out of a stone.
Min Elect Metab 1995; 20: 399-409.
10. Doyle IR, Marshall VR, Ryall RL. Calcium oxalate crystal matrix extract: the
most potent macromolecular inhibitor of calcium oxalate crystallization yet
tested in undiluted urine in vitro. Urol Res 1995; 23: S3-62.
II. Stapleton AMF, Seymour AE, Brennan JS, Doyle IR, Marshall VR, Ryall RL.
Immunohistochemical distribution and quantification of crystal matrix protein.
Kidney Int 1993; 44: 817-24.
12. Bohlen P, Stein S, Dairman W, Udenfriend S. Fluorometric assay of proteins in
the nanogram ranges. Arch Biochem Biophys 1993; 155: 213-20.
13. Castell JV, Cervera M, Marco R. A convenient micromethod for the assay of
primary amines and proteins with f1uorescamine. A re-examination of the
conditions of reaction. Anal Biochem 1979; 99: 379-91.
14. Laemmli UK. Cleavage of structural proteins during the assembly of the head
of bacteriophage H. Nature (London) 1970; n7: 680-5.
15. Heukeshoven J, Dernick R. Simplified method for silver staining of proteins in
polyacrylamide gels and the mechanism ofsilver staining. Electrophoresis 1985;
6: 103-12.
16. Dawnay ABStJ, Thornley C, Cattell WR. An improved radioimmunoassay for
urinary Tamm-Horsfall glycoprotein. Biochem J 1982; 206: 461-S.
17. Wiggins RC. Uromucoid (Tamm-Horsfall glycoprotein) forms different
polymeric arrangements on a filter surface under different physicochemical
conditions. Clin Chim Acta 1987; 162: 329-40.
18. Ryall RL, Hibberd CM, Marshall YR. A method for measuring inhibitory
activity in whole urine. Urol Res 1985; 3: 285-9.
19. Edyvane KA, Hibberd CM, Harnett RM, Marshall YR. Ryall RL.
Macromolecules inhibit calcium oxalate crystal growth and aggregation in
whole human urine. Clin Chim Acta 1987; 167: 329-38.
20. Edyvane KA, Ryall RL, Mazzachi RD, Marshall YR. The effect of serum on the
crystallization of calcium oxalate in whole human urine: inhibition disguised as
apparent promotion. Urol Res 1987; 15: 87-92.
21. Ryall RL. Harnett RM, Hibberd CM, Edyvane KA, Marshall YR. Effects of
chondroitin sulphate, human serum albumin, and Tamm-Horsfall mucoprotein
on calcium oxalate crystallization in undiluted urine. Urol Res 1991; 19: 181~.
22. Rose GA. Sulaiman S. Tamm-Horsfall mucoprotein promotes calcium oxalate
crystal formation in urine: quantitative studies. J Urol 1982; 127: 177-9.
23. Grover PK, Ryall RL, Marshall YR. Does Tamm-Horsfall mucoprotein inhibit
or promote calcium oxalate crystallization in human urine! Clin Chim Acta
1990; 190: 223-38.
24. Bezeaud A, Guillin MC. Quantitation of prothrombin activation products in
human urine. BrJ Haematol 1984; 58: 597-606.
25. Soriano-Garcia M, Padmanabhan K, de-Vos AM, Tulinsky A. The Ca'+ ion and
membrane binding structure of the Gla domain of Ca-prothrombin fragment I.
Biochemistry 1992; 31: 255~6.
26. Nakagawa Y, Abram V, Kezdy FJ, Kaiser ET. Coe FL. Purification and
characterization of the principal inhibitor of calcium oxalate monohydrate
crystal growth in human urine. J Bioi Chem 1983; 258: 1259~OO.
27. Shiraga H, Min W, VanDusen WJ, et al. Inhibition of calcium oxalate crystal
growth in vitro by uropontin: another member of the aspartic acid-rich protein
superfamily. Proc Natl Acad Sci USA 1992; 89: 426-30.
28. Atmani F, Lacour B, Strecker G, Parvy P, Drueke T, Daudon M. Molecular
characteristics of uronic-acid-rich protein. a strong inhibitor of calcium oxalate
crystallization in vitro. Biochem Biophys Res Commun 1993; 191: 115~5.