British Journal of Rheumatology 1997;36:990–995
RELATIONSHIP BETWEEN URINARY EXCRETION OF MODIFIED NUCLEOSIDES
AND RHEUMATOID ARTHRITIS PROCESS
J. G. TEBIB, C. REYNAUD,* J. P. CEDOZ, M. C. LETROUBLON and A. NIVELEAU*
CHLS, 69310 Pierre Bénite and *CNRS URA 1459 , Institut Pasteur, 69365 Lyon, France
SUMMARY
The levels of the five methylated nucleosides pseudouridine (C-Urd), 1-methyladenosine (1-MeAdo), 4 acetylcytidine (4-AcCyd),
1 methylinosine (1-MeIno) and 7 methylguanosine (7-MeGuo) resulting from RNA degradation were examined in the urine of
rheumatoid arthritis (RA) patients. Of these five, 1-MeAdo and C-Urd were correlated with the active phase of the disease,
while two others (4-AcCyd and 1-MeIno), which require further evaluation, appeared to be linked to the prognosis of the disease.
As RNA turnover is closely associated with cell proliferation, including that of lymphocytes in RA, there may be a hitherto
unsuspected benefit in measuring 1-MeAdo and C-Urd as biochemical markers of RA disease activity.
K : Rheumatoid arthritis, Disease activity, RNA catabolism.
T pathophysiology of rheumatoid arthritis (RA) is
still not well understood, but the process is characterized by the proliferation of immunocompetent cells
within the synovial membrane leading to the pannus.
This structure is potentially devastating for the joint
cartilage in which the production of numerous toxic
mediators occurs [1]. The determination of mediators
like cytokines, degradative enzymes and collagen
catabolites has been carried out, but the results of these
studies provided only a partial estimation of the RA
process [2].
Evaluation of the magnitude of lymphocyte proliferation could provide a relatively simple procedure for
assessing disease activity. It is well known that when
cell proliferation takes place, certain genes have to be
transcribed into RNA which, like all biological
molecules, possesses a turnover time dependent on the
efficiency of the protein synthesis machinery of the cell
[3]. The fidelity of translation, however, is itself affected
by the post-transcriptional modifications undergone by
the mRNA. For instance, the capping of mRNAs adds
a 7-methylguanosine (7-MeGuo) at the 5' end of the
transcript by a 5':5' link, thereby allowing the specific
binding of a protein which facilitates initiation of
mRNA translation. If 7-MeGuo is found primarily in
mRNA, other modified nucleotides (MN) are more
prevalent in rRNA and sRNA. For example, it has
been shown that pseudouridine (C-Urd) is found in
tRNA or rRNA, while N2, N2-dimethylguanosine is
primarily found in rRNA [4, 5]. When RNAs are
degraded to nucleosides and finally to bases, the four
normal bases are reutilized by the salvage pathway
for de novo nucleotide synthesis. In the case of MN,
they cannot be reused for the de novo synthesis and
are directly eliminated in urine. It was, therefore,
reasonable to undertake measurements of their urinary
concentration in order to estimate total RNA turnover.
Further, the concentration of specific MN should
permit us to evaluate the turnover of each of the
different RNA species [6].
In cancer, these MN are excreted in higher amounts
in affected patients than in healthy subjects and they
are considered to be markers of severity [7] or relapse
[8]. However, these estimations are difficult to perform
routinely because of the need for high-performance
chromatographic (HPLC) methods.
Since the pioneering work of Beiser et al. [9],
immunological techniques have been demonstrated to
be as accurate as and more convenient than HPLC for
a clinical approach [10]. Furthermore, these assays
permit more than one MN to be monitored for a given
neoplasm since several specific antibodies are available.
Consequently, they allow a higher level of confidence
to be reached in diagnosis or follow-up [11]. We present
here the preliminary results obtained when the five MN
were measured in the urine of 31 RA patients during
a 12 month follow-up.
C-Urd, 1-methyladenosine (1-MeAdo), 4-acetylcytidine (4-AcCyd) and 1-methylinosine (1-MeIno) are
known to stem from rRNA and tRNA, while mRNA
production is partially related to 7-MeGuo. The
excretion of four MN was significantly increased in RA
patients when compared with the excretion measured in
healthy subjects. Interestingly, the respective concentration of each MN was different from those observed
for cancer or haemopathies. Correlations with RA
parameters seem to indicate that C-Urd and 1-MeAdo
could be used as markers of RA activity, while the
determination of 4-AcCyd, and to a lesser extent
1-MeIno, would appear to be linked to the RA
autoimmune process.
MATERIALS AND METHODS
Patients
In order to determine whether a simple morning
sample of urine could reflect the 24 h production of
each of the MN, 10 patients suffering from RA
Submitted 1 November 1995; revised version accepted 21 February
1997.
Correspondence to: J. G. Tebib, Service de rhumatologie, Centre
Hospitalier Lyon Sud, 69310 Pierre Bénite, France.
= 1997 British Society for Rheumatology
990
TEBIB ET AL.: MODIFIED NUCLEOSIDES IN RA
hospitalized for a check-up of their disease were asked
to participate in a nyctemeral cycle study of MN
excretion during 3 days. No change in their treatment
was allowed during this period of time. A first aliquot
was immediately taken from a freshly collected uring
sample at 8.00 a.m. Two other samplings were
performed at 4.00 p.m. and 10.00 p.m., while the 24 h
urine collection was saved at 4°C until the day after the
last aliquot was collected.
In a second step, 31 patients suffering from RA were
successively examined as out-patients during a 12
month period to assess the significance of these
determinations as biological markers of the RA
process. Neither renal insufficiency nor neoplasia were
detected during the time of the survey. The mean age
of patients was 57 2 9 yr. The sex ratio was 3 women:1
man. All the patients were receiving a long-acting drug;
five patients were on an i.m. gold regimen, seven
patients were on other sulphydryl compounds,
sulphasalazine was provided to seven others, while 12
patients took methotrexate distributed weekly (7.5–
15 mg). Twenty-eight of these patients did not take
any steroids, whilst two patients had a daily dose
of 7–10 mg of prednisone. The last patient needed
more elevated amounts of steroids (steroid pulse of
15 mg/kg) because of a flare-up of his disease (samples
were collected 1 day before the pulse), but none of the
31 patients was allowed to change their DMARD
regimen during the follow-up period except when a
flare-up occurred.
The degradation of joints was rather moderate since
only one patient was graded IV in the radiological
Steinbrocker index. A total of 25% of patients
presented as grade I, while half of them were classified
as grade II and grade III was retained for the remaining
quarter of patients.
Both clinical and biological assessments were
performed on every patient at each visit. The Lee index
was used to quantify the patient’s abilities along with
morning stiffness. The swollen joint counts and the
Ritchie index were used specifically to assess the
inflammatory activity. Over this period, the erythrocyte
sedimentation rate (ESR) according to the Westergren
method and C-reactive protein (CRP) levels (nephelometric technique, Behring) were used as biological
markers of disease activity. Rheumatoid factors (RF)
were quantified by the nephelometric method (Behring)
and results are given in IU/ml. Active disease was then
defined on the basis of three or more of these four
criteria: morning stiffness (MS) lasting q45 min, a
swollen joint count and a Ritchie index grading of q3
and 10, respectively, ESR q27 mm/h. This definition
of activity corresponds to the one proposed by the
Food and Drug Administration (FDA) for clinical
evaluation of anti-rheumatic drugs, except for the
tender joint count which was replaced by the Ritchie
index.
All 31 patients were seen once, in addition, five were
seen five more times, three were seen two more times
and two were seen one more time during the 12 month
period, so that in all 64 clinical and biological assess-
991
ments were available. The five patients seen six times
were followed over a 10–12 month period which also
permitted a longitudinal study to be carried out.
Morning urinary samples were obtained together
with blood samples at each visit. The 64 urine samples
from the total population of RA were then aliquoted,
centrifuged and stored at −50°C until determination.
The control population was chosen to present the
same average age and gender ratio as the RA
population. Twenty-seven individuals without a history
of inflammatory disease provided early morning
urine samples. Further, no statistical differences in
the urine creatinine level were found between the
two populations.
Urinary determination of the modified nucleosides
The production and characterization of monoclonal
antibodies, as well as the immunoenzymatic technique, have been described in detail elsewhere [12].
Briefly, Balb-C mice were immunized by injection of
the MN (which is not immunogenic alone) coupled to
bovine serum albumin (BSA). Monoclonal antibody
production was performed according to the hybridoma
technique. The concentration of MN was determined
by a solid-phase ELISA competition procedure.
Anti-mouse immunoglobulins conjugated to peroxidase were used as second antibodies. Results are
expressed as the ratio between the concentration of
MN and urine creatinine (Jaffé technique) in order to
take into account potential variations of glomerular
filtration (nmol/mmol).
Statistical analysis
The comparative analysis of the results obtained on
RA patients and controls was performed using the
non-parametric test (Kruskal–Wallis test or Mann–
Whitney U-test) and variance analysis F-test. A value
of P Q 0.05 was accepted as significant. The regression
analysis allowed any correlation between the various
parameters to be measured, but the level was set to 0.01
to reduce the major part of the type II error risk found
in this particular statistical analysis [13].
RESULTS
Determination of the convenient time for sampling
For each MN, determinations performed on samples
collected at 8.00 a.m., 4.00 p.m. and 10.00 p.m. were
compared with those obtained at the 24 h sample in a
population of the 10 patients during three consecutive
days (one comparison per day) using a variance
analysis for small samples (Kruskal–Wallis test). No
statistical differences were found between the results
obtained on the morning sample and the others
(Table I). For convenience, then, 8.00 a.m. was chosen
as the time for collecting the daily sample.
MN values in RA patients
MN measurements performed on urine samples
from patients with various types of solid cancers or
haemopathies already published [10] were used as
positive controls (Table II). The levels of the differ-
992
BRITISH JOURNAL OF RHEUMATOLOGY VOL. 36 NO. 9
TABLE I
8.00 a.m.
4.00 p.m.
10.00 p.m.
24 h
C-Urd
1-MeAdo
1-MeIno
4-AcCyd
7-MeGuo
59.6 2 20.1
74.7 2 36.0
72 2 64.5
66.8 2 36.5
NS P = 0.66
3.41 2 2.02
3.14 2 1.8
2.82 2 0.81
3.16 2 1.18
NS P = 0.88
3.92 2 1.83
3.54 2 1.12
3.5 2 1.2
3.49 2 1.6
NS P = 0.926
3.45 2 2.56
4.24 2 3.05
2.95 2 1.80
2.33 2 1.0
NS P = 0.526
0.19 2 0.07
0.21 2 0.08
0.22 2 0.13
0.20 2 0.11
NS P = 0.92
Urinary samples were collected at different times of the day from 10 women who suffered from RA during 3 days. 8.00 a.m., 4.00 p.m. and
10 p.m. were the times chosen for the different samplings. The determination of the five MN at each time for the 10 patients was compared
to an aliquot collected on the 24 h diuresis of the same day. A Kruskal–Wallis non-parametric test was used for this statistical analysis. Results
of day 2 are presented in the table. They were similar at day 1 and 3, and did not show any statistical difference between the different time
of sampling and the 24 h aliquot for every MN. Thus, the morning sample was chosen for convenience during the rest of the study.
ent MN in RA patients are of the same order of
magnitude as those found in cancer patients. Statistical
comparison between RA patients and cancer patients
was not carried out because of the discrepancy
observed in age, renal dysfunction and cancer state.
MN determinations in solid cancer patients, for
instance, were performed at different times of the
progression of the disease without taking into account
the treatment or the tumour burden. This largely
explains the large variance encountered in these
populations and the moderate interest in statistically
comparing the MN level with the RA population which
was strictly defined like the cross-matched control
population. In this latter comparison, the MN urinary
concentrations were significantly increased in the RA
population, except for 7-MeGuo and C-Urd. However,
.. values were also larger in the RA population than
in the controls, suggesting a weak specificity of the
dosages or the presence of subpopulations in RA
patients. This latter possibility prompted us to examine
these MN determinations according to criteria like
sex, long-acting drug intake and disease activity.
Only the last factor was discriminatory according
to the FDA disease activity score.
When the RA population was subdivided into
inactive (15 samples) and active groups (16 samples)
according to the disease activity criteria (Table III),
.. values were not more significantly elevated than
those of ESR or CRP (as verified by the variance
comparison, F-test). Using this partition, 4-AcCyd and
1-MeIno levels were significantly elevated in both the
active and inactive subpopulation (P Q 0.0001/0.003
and P Q 0.0001/0.01, respectively). In the active group,
C-Urd and 1-MeAdo concentrations were also
significantly increased not only in comparison with the
normal group (P Q 0.02 and P Q 0.008, respectively),
but also with the inactive population where the C-Urd
level was significantly increased (P Q 0.01) while the
1-MeIno level was just above the significance threshold
(P Q 0.06). The inactive group showed no significant
increase in the level of these two last catabolites in
comparison with the results obtained in the control
population. When the comparison was performed
using Student’s t-test on the whole population of 64
measurements (data not shown), these results were
confirmed and amplified for C-Urd and 1-MeIno
(P Q 0.001 and P Q 0.0001, respectively) between the
active group (44 measurements) and the inactive group
(20 measurements). 7-MeGuo did not show any
significant difference between the two RA patient
TABLE II
Age
Sex ratio
Urine creat.
DD
7-MeGuo
C-Urd
1-MeIno
1-MeAdo
4-AcCyd
RA
(n = 31)
P
Control
(n = 27)
Breast cancer
(n = 39)
Ovarian cancer
(n = 19)
Acute
myeloid
leukaemia
(n = 43)
57 2 9
3/1
5.24 2 3.86
6.5 2 1.2
0.36 2 1.1
48.1 2 16.3
2.7 2 1.2
2.9 2 1.8
3.6 2 2.8
NS
NS
NS
—
NS
NS
0.002
0.03
0.0007
54 2 8
3.2/1
7.83 2 6.9
—
0.103 2 0.06
42.3 2 10
1.9 2 0.4
2.1 2 0.7
1.6 2 0.6
—
—
—
—
0.15 2 0.21
38.7 2 26.2
2.9 2 2.2
3.6 2 2.9
4.0 2 6.0
—
—
—
—
0.06 2 0.07
35.5 2 33.1
2.2 2 1.2
2.1 2 1.2
2.48 2 1.7
—
—
—
—
0.31 2 0.29
68 2 56
4.1 2 4.1
3.2 2 2.8
4.1 2 6
Urinary samples were collected from the 31 RA patients and were tested for the determination of five MN and compared with the results
obtained in a control population cross-matched for age and sex ratio. Student’s t-test was used. 1-MeIno, 1-MeAdo and 4-AcCyd were
significantly increased in the RA population (P Q 0.03 to P Q 0.0007). Since the 7-MeGuo level appears higher in RA patients, the difference
was not significant owing to the dispersion of the data. The same measurements were also performed in patients affected by different cancers
[10] and haemopathies (personal results). The level of determination was of a similar magnitude when RA samples were compared to those
found in neoplastic pathologies. DD, disease duration.
835974 RHEUM 36/9 Issue — MS 2033
993
TEBIB ET AL.: MODIFIED NUCLEOSIDES IN RA
TABLE III
Inactive RA
(n = 15)
59 2 11
7.8 2 5.1
14 2 13
4.6 2 3.1
3.1 2 2.2
19 2 10
0.13 2 0.03
41 2 11
2.6 2 0.8
2.4 2 0.8
321
Age
DD
MS
Ritchie
NSJ
ESR
7-MeGuo
C-Urd
1-MeIno
1-MeAdo
4-AcCyd
Control
(n = 27)
P
Active RA
(n = 16)
53 2 14
–
–
–
–
–
0.103 2 0.06
42 2 10
1.9 2 0.4
2.1 2 0.7
1.6 2 0.6
NS
NS
0.002
0.002
0.001
0.002
NS
0.02
0.003
0.008
0.0001
55 2 8
6.1 2 5
99 2 102
18 2 7
14.3 2 4.5
62 2 24
0.47 2 1.54
54 2 16
2.8 2 1.4
3.8 2 1.8
4.1 2 2.1
P
NS
NS
NS
0.01
NS
0.0001
Clinical and biological indices made at each consultation were used to assess the level of activity of the RA process in the 31 RA patients.
In 15 patients, the RA state was considered as inactive, while in 16 other cases it was classified as active. The comparison was then carried
out between these two populations and controls for the five MN using the U-test of Mann and Whitney. On this basis, inactive RA was different
only by the level of 4-AcCyd and 1-MeIno (P Q 0.0001 and P Q 0.01), while C-Urd and especially 1-MeAdo allowed the active population to
be differentiated from the control (P Q 0.02 and P Q 0.008). The C-Urd level was also significantly increased in active compared to inactive
RA (P Q 0.01). DD, disease duration; MS, morning stiffness; NSJ, number of swollen joints; ESR, erythrocyte sedimentation rate.
groups and the control population (Table III).
However, its insignificant increase in active disease
along with the large .. (1.54 vs. 0.06) may be due to
the very high levels found in some of the patients of this
group, while in the non-active group the mean values
were of the same magnitude as those in the control
group.
Correlation with the different clinical and biological
parameters of the disease
Table IV presents the matrix of correlations
determined between various markers in the 31 RA
patients. Results are given as the mean of the
correlation coefficients r. C-Urd and 1-MeAdo were
correlated with the RA activity parameters (MS,
Ritchie index, ESR, P between 0.01 and 0.001). The
magnitude of the correlation between these urinary
markers and clinical indices of activity (MS, Ritchie,
number of swollen joints) was usually less strong than
that between these same markers and the classical
biological ones (ESR and CRP, P Q 0.001). However,
when correlations were done on the 64 samples, these
differences were abolished (data not shown). In
addition, C-Urd and 1-MeAdo were the markers most
strongly correlated to the Lee index, which is an index
which combines disease activity and disease progression (P Q 0.001). 4-AcCyd and 1-MeIno were
strongly linked to the duration of disease and RF level
(P between 0.01 and 0.001). 7-MeGuo did not correlate
with any of the above indices.
Follow-up study
Five patients were followed during a 12-month
period. Among them, two subjects endured significant
flare-ups of the disease (Fig. 1). In these two patients,
1-MeAdo and to a lesser extent C-Urd (data not
shown) variations followed the disease activity
(represented by the Lee index on the figures). 4-AcCyd,
1-MeIno and 7-MeGuo determinations did not present
any correlation with any of the parameters measuring
TABLE IV
DD
MS
Lee
Ritchie
ESR
CRP
Latex
7-MeGuo
C-Urd
1-MeIno
1-MeAdo
4-AcCyd
DD
MS
Lee
Ritchie
ESR
CRP
Latex
7-MeGuo
C-Urd
(((
0.574
(((
(((
(((
0.446
(((
0.449
0.69
(((
0.598
0.748
0.663
0.753
0.834
(((
(((
0.615
(((
0.526
(((
0.749
0.608
0.639
0.414
(((
0.754
(((
0.561
(((
0.760
0.576
(((
(((
0.485
(((
0.391
(((
0.802
(((
(((
0.435
(((
0.524
(((
(((
(((
0.463
(((
0.613
(((
(((
0.665
0.837
0.478
0.885
(((
(((
(((
(((
0.683
0.721
0.590
1-MeIno 1-MeAdo
0.605
0.893
0.442
The correlations were carried out between the five MN and the different clinical and biological parameters of the 31 RA patients. C-Urd
and 1-MeAdo were rather strongly correlated to disease activity markers (MS, Ritchie index, ESR and CRP), while 1-MeIno and 4-AcCyd were
rather linked to disease characteristics like disease duration or rheumatoid factors. P Q 0.05 if r q 0.38; P Q 0.01 if r q 0.448; P Q 0.005 if
r q 0.501; P Q 0.001 if r q 0.57. ***No correlation.
994
BRITISH JOURNAL OF RHEUMATOLOGY VOL. 36 NO. 9
F. 1.—Five patients were followed during a 12 month period (see
Materials and methods). Among them, two presented a flare-up of
the disease (see the Lee index curve on patients 1 and 3). 1-MeAdo
and to a lesser extent C-Urd (data not shown) levels were strongly
linked to the degree of disease activity since no parallel could really
be observed between the levels of 1-MeAdo in patient 1 and patient
3 and the activity state. The lines represent the extent of the normal
value of 1-MeAdo as defined in the control population. (see Table
III control). Note that the 1-MeAdo values tend to be normal when
the flare-up is controlled. In contrast, 4-AcCyd did not follow the Lee
index curve and remains largely above the normal value of this
metabolite. The hatched rectangle below patients 1 and 3 represents
the duration of the flare up.
disease activity, RF production or disease disability. In
the three other patients, similar results were found, but
were less meaningful, most probably because the
disease activity was less intense.
DISCUSSION
The pathophysiological process of RA is not well
understood, but the various mechanisms of activation
of the immune system should proceed, at least in part,
by an overproduction of mediators [1]. In this study, we
hypothesized that this activation could also accelerate
the turnover of RNA, thereby influencing the protein
synthesis machinery. Our preliminary findings are in
agreement with this possibility since the excretion of
several MN, as measured by our ELISA procedure,
was significantly increased in RA patients when
compared with controls, except for 7-MeGuo. Further,
MN levels in RA patients deserve to be compared with
the results obtained in neoplasia. Indeed, in carcinomas, the urinary excretion reaches similar orders of
magnitude (Table I; [10, 11]). In haemopathies, C-Urd,
1-MeIno [14], 4-AcCyd and 7-MeGuo (J. G. Tebib,
unpublished results; Table I) levels are twice as high as
those measured in RA patients. However, when the RA
disease is active, this difference decreases, especially for
C-Urd and for 4-AcCyd. Of interest is the increase in
1-MeAdo excretion which is strongly increased during
flare-up, contrary to what is observed in various solid
cancers or haemopathies ([10]; J. G. Tebib, unpublished results). These comparisons deserve two
comments. First, nucleotide metabolism in the RA
process is high since the production of several MN is
significantly increased when compared with that
observed in controls. Furthermore, the respective
concentrations of some MN in RA were found to be
as high as or even higher than those detected in
neoplasia. Second, the pattern of MN excretion is
different between these two latter populations,
suggesting that there may be a difference between
neoplastic dedifferentiation and the autoimmune
activation which could be detected at the level of
nucleotide metabolism. Moreover, when these assays
were performed on 27 patients with spondyloarthropathies, only 4-AcCyd was significantly elevated
(data not shown), enhancing the hypothesis of a
possible pattern of production of MN specific for a
given disease.
The .. values of the different MN measurements
were usually larger in RA patients compared to
controls, suggesting either a weak sensitivity of these
measurements or an inhomogeneous distribution of
the RA population. The second hypothesis was tested
using the RA population separated according to
different parameters. Duration of disease and radiological degradation did not improve the magnitude of
.. as well as the intake of disease-modifying drugs,
even when different subclasses were made on the basis
of treatment or clustering of the RA population into
two classes (with or without methotrexate). The only
discriminant factor was disease activity. Indeed, both in
inactive and active RA, the .. value in the MN
determination was not significantly increased above
that obtained in the control population (except for
7-MeGuo in the active RA patient group). Moreover, using this partition, 1-MeIno and 4-AcCyd were
significantly increased in both RA subclasses, while
C-Urd and 1-MeAdo were only elevated in the active
group. Then, the larger magnitude of the .. observed
in RA patients may be explained by the variation of
disease activity from one patient to another.
To go further, the existence of correlations between
MN concentrations and events which occurred during
the course of the disease was then tested by a statistical
study performed on the 31 patients of the initial
population.
TEBIB ET AL.: MODIFIED NUCLEOSIDES IN RA
This study showed clearly that C-Urd and 1-MeAdo
were strongly linked to the parameters measuring
disease activity. They were less correlated with it than
ESR and CRP when the Ritchie index was used, but
at the same time they presented the best correlation
with the Lee index which reflects more globally the
activity state in patients having a rather limited
destruction of the joints (three-quarters of the patients
were classified as stage I or II). Moreover, of the five
patients seen at least three times during the 1 yr
follow-up, two patients presented a flare-up of the
disease (Fig. 1). In these two cases, C-Urd and
especially 1-MeAdo dramatically followed the disease
outcome. Thus, a simple ELISA test performed on
an aliquot of untreated urine could help the clinician
to appraise disease activity. C-Urd and 1-MeAdo are
the catabolites of rRNA and tRNAs, and their
increase appears meaningful as a signal of cellular
activation.
1-MeIno and 4-AcCyd appeared to be linked to
disease duration and RF level. They allowed RA
patients to be differentiated from controls regardless of
the activity state of the disease. Since these two MN
also come from the catabolism of rRNA and tRNA,
one can speculate that they could represent more
specific catabolic products found in activated T or B
cells which need to be studied. The possibility of
quantifying the accurate turnover of various RNAs
would help this type of investigation. The 7-MeGuo
assay was non-informative. Indeed, this MN was not
associated with any parameter included in this study
and its .. values in active RA patients were strongly
increased, suggesting a possible more complex linkage
with the RA process. This compound represents at least
partially the catabolism of mRNAs and one can
speculate that the RA process may use more
translation than transcription since MN from tRNA
are significantly increased, while the more specific MN
of mRNA catabolism is not.
In summary, MN determination in RA is of interest
for at least two reasons. (i) These determinations
proved that the autoimmune and inflammatory
responses present the same degree of RNA catabolism
as those depicted in some neoplastic processes. (ii) Two
of these metabolites first appear as potential markers of
disease activity and two others are linked to the RA
process since they correlate with the duration of disease
and RF production. Should these results be confirmed
in a larger population, the inclusion of a variable like
drug intake, degree of joint destruction or estimated
pannus mass, should permit both of these determinations to be used as a non-invasive method for
quantifying disease activity. 4-AcCyd and 1-MeIno
presented different patterns, suggesting a different way
of production which may be linked more tightly to the
RA process. This unexpected result deserves to be
confirmed with determinations of the differential
production of RNAs, before attempting to use these
two determinations as prognostic markers of the
disease, as some are in neoplasia.
995
A
This work was partially supported by a grant from
the research committee of the Hospice Civil de Lyon.
We would like to thank Dr G. Quash for his
constructive help in elaborating this manuscript.
R
1. Harris ED. Rheumatoid arthritis. Pathophysiology and
implications for therapy. In: Epstein FH, ed. Mechanism
of disease. N Engl J Med 1990;322:1277–89.
2. Arend WP, Dayer JM. Cytokines and cytokine inhibitors
or antagonists in rheumatoid arthritis. Arthritis Rheum
1990;33:305–15.
3. Waterlow JC. Protein turnover with special reference to
man. Q J Exp Physiol 1984;69:409–38.
4. Sander G, Topp H, Heller-Schöch G, Wieland J, Schöch
G. Ribonucleic acid turnover in man: RNA catabolites
in urine as measure for the metabolism of each of the
three major species of RNA. Clin Sci 1986;71:367–74.
5. Topp H, Sander G, Heller-Schöch G, Schöch G.
Determination of 7 methylguanine, N2,N2-dimethylguanosine and pseudouridine in ultrafiltrated serum of
healthy adults by high-performance liquid chromatography. Anal Biochem 1987;161:49–56.
6. Sander G, Topp H, Wieland J, Heller-Schöch G,
Schöch G. Possible use of urinary modified RNA
metabolites in the measurement of RNA turnover in
the human body. Hum Nutr Clin Nutr 1986;40c:
103–18.
7. Viale GL, Fondelli-Restelli A, Viale E. Methylated bases
in tRNAs of brain tumors. Tumori 1967;53:533–40.
8. Salvatore F, Colonna A, Costanzo F, Russo T, Esposito,
Cimino F. Modified nucleosides in body fluids of
tumor-bearing patients. Recent Results Cancer Res
1983;84:361–77.
9. Munns TW, Liszewski MK. Antibodies specific for
modified nucleosides: An immunological approach for
the isolation and characterization of nucleic acids. In
Research and molecular biology, 1980, vol. 24, New
York: Academic Press, pp. 109–63.
10. Reynaud C, Bruno C, Boullanger P, Grange J, Barbesti
S, Niveleau A. Monitoring of urinary excretion of
modified nucleosides in cancer patients using a set of
six monoclonal antibodies. Cancer Lett 1991;61:
255–62.
11. Müller J, Erb N, Heller-Schöch G, Lorenz H, Winkler
K, Schöch G. Multivariate analysis of urinary RNA
catabolites in malignancies: cross-sectional and longitudinal studies. Recent Results Cancer Res 1983;84:
317–30.
12. Reynaud C, Niveleau A. Methylated bases as tumor
markers: detection and quantification with a competitive enzyme-linked immunoassay. Anal Lett 1990;23:
31–45.
13. Robert C. Analyse descriptive multivariée: Application à
l’intelligence artificielle. Medecine-Sciences Flammarion,
1989.
14. Itoh K, Konno T, Sasaki T, Ishiwata S, Ishida N,
Mizugaki M. Relationship of urinary pseudouridine
and 1-methyladenosine to activity of leukemia and
lymphoma. Clin Chim Acta 1992;206:181–9.