2658 Biophysical Journal Volume 74 May 1998 2658 –2665

Time-Dependent Effects of Trimethylamine-N-Oxide/Urea on

Lactate Dehydrogenase Activity: An Unexplored Dimension of the
Adaptation Paradigm

Ilia Baskakov and D. W. Bolen

Department of Human Biological Chemistry and Genetics, University of Texas Medical Branch, Galveston, Texas 77555-1052 USA

ABSTRACT Given that enzymes in urea-rich cells are believed to be just as sensitive to urea effects as enzymes in
non-urea-rich cells, it is argued that time-dependent inactivation of enzymes by urea could become a factor of overriding
importance in the biology of urea-rich cells. Time-independent parameters (e.g. Tm, kcat, and Km) involving protein stability
and enzyme function have generally been the focus of inquiries into the efficacy of naturally occurring osmolytes like
trimethylamine-N-oxide (TMAO), to offset the deleterious effects of urea on the intracellular proteins in the urea-rich cells of
elasmobranchs. However, using urea concentrations found in urea-rich cells of elasmobranches, we have found time-
dependent effects on lactate dehydrogenase activity which indicate that TMAO plays the important biological role of slowing
urea-induced dissociation of multimeric intracellular proteins. TMAO greatly diminishes the rate of lactate dehydrogenase
dissociation and affords significant protection of the enzyme against urea-induced time-dependent inactivation. The effects
of TMAO on enzyme inactivation by urea adds a temporal dimension that is an important part of the biology of the adaptation
paradigm.

INTRODUCTION
The biology of adaptation involves the study of organisms Up to the present, we have focused on one of the two
that have adapted to environmental stresses such as ex- defining characteristics of osmolytes, namely the mecha-
tremes of temperature, dehydration, contact with high-salt nism by which osmolytes stabilize proteins thermodynam-
solutions, and even the presence of intracellular concentra- ically against denaturing stresses. Our results provide strong
tions of urea (Yancey et al., 1982). Organisms that have evidence that the unfavorable interaction of osmolytes with
adapted to these stresses appear to concentrate small organic the peptide backbone of protein is responsible for the ther-
compounds, called osmolytes (Brown and Simpson, 1972; modynamic stabilization (Wang and Bolen, 1997). But in
Stewart and Lee, 1974; Yancey et al., 1982), whose intra- the work described here, we address the second defining
cellular concentration in the range of several hundred characteristic of osmolyte action. Namely, we investigate
millimolar is believed to have two defining characteristics: the ability of the naturally occurring osmolyte trimethyl-
1) the osmolytes have the ability to stabilize cellular pro- amine N-oxide (TMAO) to offset the deleterious effects of
teins against the inactivating stress for which the osmolytes the 400 – 600 mM intracellular urea concentrations in the
were naturally selected, and 2) the osmolytes do not greatly cells of elasmobranchs.
perturb the functional activities of proteins or otherwise Many of the efforts to investigate the effects of osmolytes
upset the delicate metabolic control systems necessary to on protein function have focused on kcat, Km, and Ki param-
sustain life (Somero, 1986; Yancey et al., 1982; Yancey and eters of enzymes (Burg et al., 1996; de Meis, 1988; Gopal
Somero, 1980). These characteristics focus on the thermo- and Ahluwalia, 1993; Lin and Timasheff, 1994; Santoro et
dynamic stability and function of proteins in the face of al., 1992; Wang and Bolen, 1996; Yancey and Somero,
environmental stresses, and form the paradigm for discuss-
1979, 1980). These parameters are assumed to be time-
ing osmolyte involvement in the biology of adaptation.
independent, but the manner in which such data are col-
Nature’s use of osmolytes in the adaptation of organisms to
lected can unwittingly incorporate time-dependent effects
environmental stresses involves one of the most elusive and
into the measurements. In the course of our investigations,
long-standing problems in biology, the general problem of
we discovered time-dependent effects of both urea and
how proteins and solvent interact to produce biologically
TMAO on lactic dehydrogenase (LDH) activity that defi-
significant effects.
nitely affect evaluations of kcat and Km (Baskakov et al.; see
companion paper). If catalytic activity of an enzyme is not
maintained for a reasonable time while the enzyme is bathed
Received for publication 1 December 1997 and in final form 27 January in the milieu containing urea, TMAO, and urea/TMAO
1998.
mixtures as it would be in urea-rich cells, the issues of
Address reprint requests to Dr. D. W. Bolen, Department of Human
“protein stabilization” and apparent kcat and Km effects
Biological Chemistry and Genetics, University of Texas Medical Branch,
5.154 Medical Research Building, Galveston, TX 77555-1052. Tel.: 409- become moot as an adaptive strategy. Clearly, the question
772-0754; Fax: 409-747-4751; E-mail: wbolen@hbcg.utmb.edu. of the lifetime of proteins in the presence of the denaturing
© 1998 by the Biophysical Society stress and (protective) osmolytes has considerable bearing
0006-3495/98/05/2658/08 $2.00 on how effective osmolytes are in the biology of adaptation.
Baskakov and Bolen Time-Dependent Effects of Urea and/or TMAO on LDH Activity 2659

With the goal of better understanding the requirements Determination of the apparent order
for TMAO stabilization of proteins in the presence of urea, of reactivation
we have investigated the time-dependent effects of urea and
These experiments were performed with an LDH sample that had been
TMAO on rabbit muscle LDH stability and function. LDH incubated for 24 h with 0.8 M urea (0.2 M Tris-HCl buffer, pH 7.3) at 0°C.
is essential for a large number of organisms, and like many Different amounts of this urea-containing solution were withdrawn and
multisubunit proteins it is rather labile (Cho and Swaisgood, diluted into assay mixtures containing 5 mM pyruvate and 180 mM NADH
1973). The effect of denaturing stress and osmolyte concen- (0.2 M Tris-HCl, pH 7.3) to obtain 340 nm absorbance versus time data,
with a range of LDH concentrations from 0.0134 to 0.178 mg/ml. The first
tration on the lifetime of enzyme activity adds a different
derivatives of the absorbance versus time curves show that NADH oxida-
dimension to the discussion of the biology of adaptation and tion increases with time in the assay, reflecting a reactivation of LDH
the issue of defining the temporal response to inactivating during the course of the assay. The initial rates of LDH reactivation,
stress in the cell. observed at different enzyme concentrations, were determined from the
initial slopes of these first derivative curves and reported as the initial rates
of reactivation. These data were fitted to a linearized form of the equation,
velocity 5 kcn, as suggested by van’t Hoff, viz., log v 5 log k 1 n log c,
EXPERIMENTAL PROCEDURES where k is the rate constant, n is the reaction order, and c is the concen-
tration of LDH (Laidler, 1965). The reaction order is obtained from the
Chemicals slope of the log v versus log c plot.
Rabbit muscle lactate dehydrogenase, NADH, sodium pyruvate, bovine
serum albumin, and TMAO were purchased from Sigma; trisma base was
from Fisher; sodium chloride was from Mallinckrodt; and ultrapure urea Gel filtration
was from ICN. Further purification of urea and TMAO and determinations
of their concentrations were performed as described in an earlier paper The experiments were carried out using a Phenomenex Biosep SEC-S3000
(Wang and Bolen, 1997). The concentration of LDH was determined high-performance liquid chromatography (HPLC) gel filtration column
spectrophotometrically at 280 nm (1.13 mg/mL/OD; supplied by with dimensions of 300 3 7.80 mm. LDH samples (13.4 mg/ml) were
Worthington). incubated in gel filtration buffer (0.10 M Tris-HCl containing 0.2M NaCl,
pH 7.3) at 0°C with either 0.8 M urea alone or a mixture of 0.8 M urea plus
0.8 M TMAO. In the course of incubation, aliquots were removed, and
apparent molecular masses were evaluated using the Phenomenex gel-
Measurement of LDH activity filtration HPLC column equilibrated in the presence of either 0.8 M urea or
0.8 M urea plus 0.8 M TMAO containing gel filtration buffer at 22°C. The
LDH activity was determined by following the oxidation of NADH to Phenomenex column was calibrated in the absence of urea and urea/TMAO
NAD1 at 340 nm. The standard reaction mixture contained 2.5 mM mixture and in the presence of 0.8 M urea, using Sigma gel filtration
pyruvate, 85 mM NADH in 0.2 M Tris-HCl buffer (pH 7.3) at 24°C. molecular mass standards (MW-GF-200) containing blue dextran (2.000
kDa), b-amylase (200 kDa), alcohol dehydrogenase (150 kDa), bovine
serum albumin (66 kDa), carbonic anhydrase (29 kDa), and cytochrome c
Time-dependent activity measurements (12.4 kDa). The steel-jacketed column was operated with mechanical
injection within a fully automated BioCad SPRINT HPLC system, which
Aliquots of concentrated LDH stock solution were added to solutions allowed the elution volume to be repeatable within 60.015 ml.
containing 0.2 M Tris-HCl buffer (pH 7.3), 1 mg/ml bovine serum albumin
(BSA), and desired amounts of either urea, TMAO, or urea plus TMAO.
These solutions were incubated on ice and, with respect to time of incu- RESULTS
bation, the LDH activity was monitored using initial velocities by addition
of aliquots of the incubating solution to standard assay mixtures. The pH Inactivation of LDH by urea
values of Tris-HCl buffer solutions to be used at 0°C were adjusted to pH
7.3 at 5°C, and those to be used at 24°C were adjusted to pH 7.3 at 25°C. Fig. 1 shows data on the time-dependent inactivation of
LDH in the course of incubation with different concentra-
tions of urea. Two different protocols for testing enzyme
Reactivation of urea-inactivated LDH activity in the course of incubation can be used, one in
which the assay mixtures contained the same concentrations
Time-dependent loss of LDH activity was induced by incubation of LDH of urea as those in the incubation samples, and a second in
(13.4 mg/ml) in 0.2 M Tris-HCl buffer (pH 7.3) containing 0.8 M urea and
1 mg/ml BSA at 0°C. To initiate reactivation from the inactivating effects which the common standard assay mixture did not contain
of urea, aliquots of the urea-containing incubation mixture were taken with urea. The two cases give different absolute velocities, but by
time and diluted 10-fold into 0.2 M Tris-HCl buffer (pH 7.3), containing 1 reporting the velocity at a given incubation time divided by
mg/ml BSA held at 24°C. In the course of incubation at 24°C, aliquots the velocity at zero time of incubation, the resulting per-
from this solution were withdrawn and initial velocities were assayed in the centages of initial activity for the two protocols were found
standard reaction mixture. Two different procedures were used to initiate
reactivation. In one, aliquots of the enzyme in 0.8 M urea at 0°C were to be identical. The second protocol was experimentally
withdrawn and diluted 10-fold into Tris-HCl/BSA buffer equilibrated at simpler, and the data from this protocol are reported in
24°C as described above. In the second procedure, aliquots of enzyme were Fig. 1.
added to the Tris-HCl/BSA buffer equilibrated at 0°C, and then rapidly Upon incubation at 0°C in the absence of urea, LDH loses
(within 1 min) warmed to 24°C. No differences in the final level of LDH activity with a half-time of ;105 min. Inclusion of urea in
activity were detected with these two procedures, but the second method
allowed us to follow the initial time dependence of reactivation, and this the incubation mixture greatly accelerates the processes of
procedure, being more convenient, was used exclusively in the work inactivation, such that in 0.8 M urea, LDH is inactivated
described. with a half-time of ;40 min.
2660 Biophysical Journal Volume 74 May 1998

FIGURE 1 Time course of LDH inactivation in the presence and ab-

sence of urea. LDH (1.34 mg/ml) activity versus time of incubation at 0°C
in 0.2 M Tris-HCl buffer containing 1 mg/ml BSA, pH 7.3, without urea
(M), and with 0.2 M (l), 0.4 M (Œ), 0.6 M (F), and 0.8 M (f) urea.
Residual LDH activity is presented as a percentage of the specific activity
measured at zero time of incubation.
FIGURE 2 Kinetics of LDH reactivation. LDH (13.4 mg/ml) was incu-
bated at 0°C in 0.2 M Tris-HCl buffer containing 1 mg/ml BSA, pH 7.3, in
the presence of 0.8 M urea for 23, 48, and 96 h, at which times an aliquot
Reversibility and reactivation kinetics was removed and diluted 10-fold into the buffer alone at 0°C. Then the
sample was warmed to 24°C within 1 min, and aliquots were assayed with
To obtain more insight into the nature of the time-dependent
respect to the time of 24°C incubation. The results of the activity mea-
urea-induced LDH inactivation, the reversibility of the in- surements are shown as the percentage yield of reactivated enzyme at 24°C
activation was studied. Reversibility was tested by with- after exposure to 0.8 M urea for 23 h (f), 48 h (F), and 96 h (Œ) at 0°C.
drawing aliquots of LDH incubated for a specified period of The inset shows a second-order plot of the 96-h LDH data (Œ) given in the
several hours in 0.8 M urea at 0°C, then diluting the aliquots main figure. Also shown in the inset is a second-order plot of the reacti-
vation of LDH that had been exposed for 96 h at 0°C to 0.8 M urea, then
10-fold into buffer solution at 0°C. Because reactivation
reactivated at 24°C in the presence of 2.5 mM pyruvate and 80 mM
was found to depend significantly on temperature, with an NADH (3).
increase in the temperature from 0°C to 24°C promoting
both the rate and the yield of reactivation (data not shown),
the reactivation was initiated by warming the samples from second-order (1/[% inactive enzyme] versus time) plot in
0 to 24°C within a period of 1 min. The activity of the the inset of Fig. 2 suggests LDH reactivation may follow
enzyme in the diluted solution at 24°C was monitored with second-order kinetics. Furthermore, apparent second-order
time of incubation at 24°C, using initial velocity measure- kinetics was observed regardless of whether reactivation
ments. Fig. 2 presents data on LDH reactivation kinetics as was performed in the presence or absence of NADH and
a function of LDH exposure to 0.8 M urea at 0°C for 23, 48, pyruvate, with the presence of the coenzyme and substrate
and 96 h. We note that after 23, 48, or 96 h at 0°C, the increasing the rate of reactivation but not affecting the
residual LDH activities do not exceed 2% of initial activity apparent reaction order.
(see data in Fig. 1). As indicated by the data in Fig. 2, LDH A possible complication to the measurements is the pres-
inactivated by 0.8 M urea incubation at 0°C for $23 h can ence of bovine serum albumin (BSA) in the inactivation and
be recovered, but the level of reactivation is incomplete and reactivation solutions at concentrations 100-1000-fold
depends upon the time LDH remains in 0.8 M urea at 0°C; higher than that of LDH. BSA is commonly added to
the longer the incubation time, the lower the extent of prevent LDH loss due to adsorption to glass or loss by
reactivation. surface denaturation. To determine the concentration depen-
The inset in Fig. 2 gives a replot of 96-h reactivation data dence of LDH reactivation and to ensure that BSA does not
in the form of a second-order kinetic plot. The percentage of affect the apparent order of reactivation, we performed
inactive enzyme is evaluated from the relationship % inac- additional experiments without BSA over a range of LDH
tive 5 100(A` 2 At)/(A`), where At is the activity (DOD/ concentrations. From such data it is possible to determine
min) at time t, and A` is the activity (DOD/min) at infinite the apparent order of reactivation by using the generalized
time of reaction, here taken as 0.044 DOD/min, the activity expression log(V) 5 log(k) 1 nzlog(C), where V is the initial
after 60 min of reactivation time. The linearity of the rate of reactivation, C is LDH concentration, and n is the
Baskakov and Bolen Time-Dependent Effects of Urea and/or TMAO on LDH Activity 2661

apparent reaction order (Laidler, 1965). Varying concentra- assay mixtures, and a log (relative initial velocity) versus
tions of LDH inactivated with 0.8 M urea for 24 h were log (relative LDH concentration) as specified by the general
placed in assay mixtures containing high concentrations of expression given above is shown in Fig. 4. The final LDH
pyruvate and NADH, and the 340-nm absorbance change concentrations ranged from 0.0134 mg/ml to 0.178 mg/ml,
was monitored with time as described in the Experimental giving corresponding relative reactivation velocities cover-
Procedures. Fig. 3 A shows the results of such assays, and it ing a 250-fold change in velocity. The slope of the plot (n 5
is clear from the acceleration of absorbance change during 2.05 6 0.08) shows the reactivation to be second order, in
the time of the assay that LDH was being reactivated during agreement with the results in the inset of Fig. 2. Regardless
the course of the assay. The time derivative of these absor- of whether BSA is present (Fig. 2, inset) or absent (Fig. 4),
bance versus time curves as shown in Fig. 3 B defines the reactivation appears to be second order, indicating that the
rate of LDH reactivation at 24°C, and the initial slope of the order of the reactivation reaction does not depend upon the
derivative plot at each LDH concentration is taken as the presence of BSA.
initial rate of active LDH appearance within the assay time.
A ratio is taken of each of the initial velocities to the initial
Subunit dissociation
velocity at the lowest LDH concentration, to give the rela-
tive initial velocities for the assays. A ratio is also taken of Gel filtration experiments were performed to test whether
each LDH concentration to that of the lowest LDH concen- time-dependent LDH inactivation by urea is caused by
tration to give the relative total LDH concentrations in the dissociation of the tetrameric form of the enzyme. The gel
filtration column was calibrated with molecular mass stan-
dards, conducted in the presence and absence of 0.8 M urea,
and identical calibration plots relating molecular mass to
elution volumes were obtained. This demonstrates that the
permeation properties of the gel and the integrity of the
molecular mass standards do not change significantly in the
presence of 0.8 M urea.
Data on the evaluation of LDH apparent molecular mass
as a function of incubation time in 0.8 M urea are presented
in Fig. 5 A. With time of incubation, the relative amount of
protein corresponding to the major peak (left peak) at zero

FIGURE 4 van’t Hoff plot of the rate of appearance of reactivated LDH

as a function of LDH concentration. The rates of appearance of reactivated
FIGURE 3 LDH reactivation dependence on LDH concentration. (A) LDH were obtained from Fig. 3 B. To convert to relative initial velocities,
Representative assays as a function of LDH concentration. The LDH the rate of appearance of LDH activity at the LDH concentration in
solutions had been $98% inactivated by incubation of LDH (13.4 mg/ml question is divided by the rate of appearance of reactivated LDH obtained
in 0.2 M Tris-HCl, pH 7.3) for 24 h in 0.8 M urea at 0°C. Aliquots of this for the lowest concentration of LDH (viz. 0.0134 mg/ml). The relative LDH
incubation were diluted from 50- to 1000-fold with the buffer at 0°C, and concentration was obtained by dividing the LDH concentration in question,
then assayed in the presence of high concentrations of substrates at 24°C. by the lowest LDH concentration. The highest concentration of protein
Final LDH concentration in mg/ml, from left to right: 0.178, 0.148, 0.0893, represented is 0.178 mg/ml. A log-log (van’t Hoff) plot is presented of the
0.0446, 0.0223, and 0.0134. (B) First derivatives of the 340 nm absorbance final relative concentrations of LDH present in the assays versus the initial
versus time curves of the assays in A were obtained, and the slopes at time relative rates of appearance of active LDH. The slope of the plot (n 5
0 of the derivative plots were taken to represent the initial rates of 2.05 6 0.08) represents the apparent reaction order of the reactivation
appearance of reactivated LDH as a function of LDH concentration. process.
2662 Biophysical Journal Volume 74 May 1998

essentially disappears after 1280 min of incubation with

urea (Fig. 5 A) at 0°C. The sum of the areas of the two peaks
decreases with time, suggesting the species at the larger
elution volume has a lower absorptivity than the species
with small elution volume. The position of the left peak
does not change with time of incubation, whereas the right
peak shifts slightly with time in a direction consistent with
higher molecular mass species.
The same type of gel filtration experiment was performed
with LDH incubated in the presence of a 0.8 M TMAO:0.8
M urea mixture (Fig. 5 B). Here LDH incubated for up to
1380 min in the TMAO:urea mixture appears as a single
peak with an elution volume essentially the same as that of
the initial chromatogram in Fig. 5 A. These results suggest
that at LDH concentration as low as 13.4 mg/ml, urea causes
time-dependent changes in the quaternary structure of LDH,
whereas the addition of TMAO to the urea prevents these
changes from occurring.
To define the elution positions of the enzyme in the
highest and lowest states of association, the dependencies of
apparent molecular mass on LDH concentration in the pres-
ence and absence of 0.8 M urea were studied. In both
experiments (with and without urea), the major peak (Ve 5
8.36 ml) contained more than 99% of total absorption, and
two very minor peaks were detected (Table 1), one with an
apparent molecular mass approximately that of an octamer,
and the other with an apparent molecular mass near that of
a single subunit. The species composing the major peak has
FIGURE 5 Gel filtration of LDH incubated with 0.8 M urea or a 0.8 M an apparent molecular mass of ;100 kDa, which is between
urea:0.8 M TMAO mixture as a function of time. The apparent molecular the molecular mass of a tetramer (144 kDa) and a dimer
mass of LDH was estimated with a Phenomenex Biosep SEC-S3000 HPLC
gel filtration column in gel filtration buffer (0.10 M Tris-HCl, pH 7.3, with
(72 kDa). An apparent intermediate molecular mass is sug-
0.2 M NaCl) at 22°C. (A) The column was equilibrated with 0.8 M urea in gestive of a tetramer-dimer equilibrium that is rapid relative
the gel filtration buffer at 22°C, and elution volumes of LDH (13.4 mg/ml) to the chromatographic time scale, with an average elution
were monitored as a function of the time of incubation in 0.8 M urea at 0°C volume composed of the weighted averages of the forms in
for 5 min. (zzzzz) (1); 40 min. (2); 80 min. (3); 160 min. (4); 240 min. (5);
equilibrium. If this interpretation applies, the elution vol-
1230 min. (——) (6). (B) The column was equilibrated with a 0.8 M
urea:0.8 M TMAO mixture in 0.1 M Tris-HCl (pH 7.3) with 0.2 M NaCl ume will shift toward the dimer elution volume with de-
at 22°C, and elution volumes of LDH (13.4 mg/ml) were monitored as a creasing LDH concentration. It was found, however, that the
function of the indicated time of incubation in the urea: TMAO mixture at elution volume and the ratio of the major peak to the minor
0°C, 5 min. (zzzzz); 80 min.(- - -); 240 min. (– – –); 375 min. (——); 1380
peak corresponding to a monomer do not depend on the
min. (——)
concentrations of enzyme in the range 5.2– 670 mg/ml either
with or without urea (see Table 1). This suggests that the
time (elution volume 8.36 ml) is observed to decrease reason for deviation of the molecular mass is not due to
concomitantly with an increase in the area of the peak at the rapid equilibrium between dimer and tetramer, but appar-
larger elution volume (right peak in figure). The left peak ently arises from other effects.

TABLE 1 Elution volumes of LDH species obtained after 5 min of incubation at 0°C in the presence of the solutes indicated
Elution volume (ml) and corresponding (kDa)* of
Conditions Concentrations of LDH 1st minor peak# Major peak 2nd minor peak
No solutes 5–670 mg/ml §
7.673 6 0.005 (230) 8.442 6 0.007 (97.6) 9.640 6 0.010 (24.2)
0.8 M urea 5–670 mg/ml§ 7.50 6 0.005 (277) 8.352 6 0.009 (103) 9.315 6 0.015 (34)
0.8 M urea 1 0.8 M TMAO 13.4 mg/ml — 8.464 6 0.014 (93.1) —
*Average elution volumes corresponding to different LDH concentrations. Values of molecular masses determined from calibration plots are presented in
parentheses.
#
The two minor peaks become apparent only on the most sensitive absorbance scale (0.001 OD at 280 nm).
§
Eight different concentrations between 5 and 670 mg/ml were used.
Baskakov and Bolen Time-Dependent Effects of Urea and/or TMAO on LDH Activity 2663

The minor peak reported in Table 1 with an apparent protection of LDH from time-dependent activity loss in the
molecular mass of 24 kDa is lower than the known mass (36 presence of urea, but it is unable to completely prevent loss
kDa) of a LDH subunit. This observation suggests that of activity in the concentration range and ratios of 3:2 or 2:1
column sorption effects may be responsible for the anoma- urea:TMAO found in sharks and rays (Yancey and Somero,
lously low molecular masses. The fact that apparent molec- 1979).
ular masses of the three detectable chromatographic species A quantitative assessment of how the half-time for LDH
in the presence of urea are higher than the corresponding inactivation varies with urea, TMAO, and 1:1 and 2:1 ratios
peaks in the absence of urea (see Table 1) could be due to of urea:TMAO is given in Fig. 7. At both 1:1 and 2:1 ratios
0.8 M urea partially abolishing the adsorption effects and/or of urea:TMAO, as urea concentration increases TMAO is
inducing swelling of the protein species. A swelling effect seen to provide greater and greater improvement in staving
caused by urea on the native forms of proteins has been off inactivation, compared to activity loss that would occur
observed by means of size exclusion chromatography (Cor- if TMAO were not present.
bett and Roche, 1984).
DISCUSSION
Effect of TMAO on urea-induced time-dependent
Urea-rich cells such as those in elasmobranches and kidney
inactivation of LDH
pose the problem that intracellular enzymes must reside in
At issue is the question of how TMAO affects the time- the presence of the denaturant (urea), yet must also maintain
dependent inactivation of LDH by urea. Fig. 6 shows results functional activity. Enzymes in urea-rich cells are believed
of the same type of experiment as shown in Fig. 1, but with to be just as sensitive to the deleterious effects of urea as
incubations carried out in the presence of TMAO alone and those occurring in non-urea-containing cells (Yancey et al.,
in mixtures of urea:TMAO. Upon incubation of LDH in 0.6 1982; Yancey and Somero, 1978). What protects the en-
M TMAO, it is found that the enzyme is inactivated by a zymes in urea-rich cells and enables them to function is the
measurable but relatively small extent, with a half-time of additional presence of organic osmolytes such as TMAO in
inactivation that is about twofold less than that of the elasmobranches and glycerolphosphocholine in kidney cells
control. Urea (0.6 M) inactivates the enzyme at a rate that is (Bagnasco et al., 1986; Garcia-Perez and Burg, 1990;
300-fold faster than that of the control, but when 0.3 M Yancey, 1985; Yancey and Somero, 1978). Much of the
TMAO is present with 0.6 M urea, the half-time of the emphasis in the biology of adaptation of urea-rich cells
inactivation rate is only 60-fold faster. If TMAO concen- has been on the magnitude of the effects of urea and of
tration is increased to give a mixture of 0.6M urea:0.6M
TMAO, the half-time of inactivation is only ;10-fold faster
than that of the control. Clearly, TMAO provides significant

FIGURE 7 Dependence of the half-time of LDH activity loss as a

function of urea and/or TMAO concentration. The half-times of activity
loss were evaluated from incubations conducted at various concentrations
FIGURE 6 Time course of LDH activity loss during incubation with of urea, TMAO, and urea:TMAO mixtures from data such as those pre-
urea, TMAO, or urea-TMAO mixtures. LDH (1.34 mg/ml) was incubated sented in Fig. 6. Represented are incubations of LDH (1.34 mg/ml) in 0.2
in 0.2 M Tris-HCl buffer containing 1 mg/ml BSA (pH 7.3) at 0°C without M Tris-HCl containing 1 mg/ml BSA (pH 7.3), carried out in the absence
solute (M), in the presence of 0.6 M TMAO (3), 0.6 M urea, and 0.6 M of solutes (M), in the presence of TMAO (3), 1:1 urea:TMAO (Œ), 2:1
TMAO (Œ), 0.6 M urea and 0.3 M TMAO (F), 0.6 M urea (f). Residual urea:TMAO (F), and urea alone (f). The concentration scale represents
LDH activity is presented as a percentage of the specific activity measured the solute concentration or, in the case of urea:TMAO mixtures, the
at zero time of incubation. concentration of the urea component.
2664 Biophysical Journal Volume 74 May 1998

counteracting osmolyte (e.g., TMAO) on the stability and slightly with time of incubation. TMAO in a 1:1 ratio with
function of enzymes (Burg et al., 1996; de Meis, 1988; urea does not give complete protection of LDH from inac-
Gopal and Ahluwalia, 1993; Lin and Timasheff, 1994; tivation after 24 h incubation in the mixture (Figs. 6 and 7),
Santoro et al., 1992; Wang and Bolen, 1996; Yancey and so the absence of lower molecular mass species (Fig. 5 B) at
Somero, 1979, 1980). Accordingly, the parameters gener- .24 h incubation indicates that loss of activity does not
ally sought are changes in Tm as a function of urea and/or have to result entirely from LDH dissociation.
TMAO and changes in kcat and Km of enzymes as a function Although the effects are exceedingly slow, TMAO itself
of these solutes. Such parameters are truly time-independent causes LDH inactivation, decreasing the half-time of inac-
parameters if the experiments for determining Tm, kcat, and tivation in the absence of solutes from ;60 days to ;30
Km are performed in a manner that excludes time-dependent days at 0°C (see Fig. 6). Nevertheless, TMAO is quite
effects (Hand and Somero, 1982; Withycombe et al., 1965). effective in extending the half-time of the enzyme in the
However, in many reports, precautions to exclude time- presence of urea, decreasing the urea-induced rate of loss by
dependent effects have not been implemented, and the pos- fivefold in 0.6 M urea:0.3 M TMAO and by ;30-fold in 0.6
sibility exists that such evaluations of the above-mentioned M urea:0.6 M TMAO (Fig. 6). Thus it appears that TMAO
parameters may not entirely reflect the interpretation of- has the ability to greatly diminish time-dependent loss of
fered for them. Given that enzymes in the urea-rich cells of LDH activity, presumably by preventing urea-induced dis-
sharks and rays are continuously bathed in the urea:TMAO sociation of LDH.
milieu, it is important to evaluate time-dependent effects of TMAO has also been reported to decrease time-depen-
these solutes on enzyme function, because such dependen- dent loss of phosphofructokinase activity due to cold-
cies are an integral part of adaptation phenomena. induced tetramer dissociation, but it does not protect the
As shown in Fig. 1, LDH held at 0°C in the absence of enzyme from inactivation due to urea-induced dissociation
urea loses activity with a half-time of ;60 days, but in 0.6 (Hand and Somero, 1982). There is an important biological
M urea, a concentration reported in the cells of rays (Forster reason why this enzyme may be an exception. Because
phosphofructokinase activity regulates glycolytic flux, it
and Goldstein, 1976), the loss of activity increases by sev-
has been hypothesized that the urea-induced dissociation of
eral orders of magnitude and occurs with a half-time of ;3
tetrameric enzyme prevents build-up of damaging acidic
h. Antidiuretic rat kidney has been reported to have urea
conditions, thereby playing an important role in the urea-
concentrations of 1.5 M (Garcia-Perez and Burg, 1990), and
rich cellular environment occurring in hibernating animals
5 M urea has been reported in desert mice under dehydra-
(Hand and Somero, 1982).
tion stress (MacMillen and Lee, 1967). Given the sensitivity
Time-dependent processes have not been emphasized in
of activity loss of LDH in urea, time-dependent effects on
the context of the biology of adaptation, but their impor-
activity must be an exceedingly important issue in under-
tance becomes evident when effects of urea on multisubunit
standing the biochemistry of urea-rich cells.
proteins are considered. Urea-induced dissociation of mul-
Fig. 2 illustrates that LDH inactivated in the presence of tisubunit proteins can occur in the presence of very low
0.8 M urea at 0°C can be reactivated at 24°C, but revers- concentrations of urea (Hand and Somero, 1982). If it is true
ibility of reactivation is dependent on the length of time of that enzymes in urea-rich cells are just as sensitive to urea
inactivation at 0°C. These data also suggest that reactivation effects as enzymes in non-urea-rich cells (Yancey et al.,
is a second-order process, and this implies that inactivation 1982), the importance of time-dependent effects becomes
by urea involves the dissociation of LDH. More than 98% of inescapable. In fact, with some enzymes it could be more
LDH activity is lost when it is incubated with 0.8 M urea at important than the counteracting effect on kcat and Km.
0°C for 24 h, but when the temperature is shifted to 24°C, Enzyme turnover is a key factor in metabolic control, and
reactivation occurs rapidly enough to see the velocity in- loss of enzyme activity as an important part of the turnover
crease during the course of the assay for LDH activity phenomenon will affect the survivability of cells stressed by
(Fig. 3 A). The time derivatives of the assays (Fig. 3 B) can urea. From the results with LDH, it is indeed likely that the
be used to evaluate the rate of reactivation as a function of ability of counteracting osmolytes like TMAO to slow down
LDH concentration, and these data permit an evaluation of or prevent dissociation of proteins in urea-rich cells is an
the reactivation reaction as second order by using the van’t essential part of the protection and counteraction it offers
Hoff (log-log) plot (Fig. 4) (Laidler, 1965). Again, the against urea-induced inactivation.
results suggest that urea has the effect of dissociating LDH
and that reassociation of dimers to active tetramers is rate
determining. Supported by National Institutes of Health grant GM49760.
Different molecular mass species can be separated with
time of LDH incubation in 0.8 M urea at 0°C (Fig. 5 A), and
it is clear that urea causes dissociation of LDH, presumably REFERENCES
to monomeric species. It is also clear that TMAO prevents
Bagnasco, S., R. Balaban, H. M. Fales, Y.-M. Yang, and M. Burg. 1986.
urea from dissociating LDH (Fig. 5 B), although the tet- Predominant osmotically active organic solutes in rat and rabbit renal
rameric species appears to change its elution properties medullas. J. Biol. Chem. 261:5872–5877.
Baskakov and Bolen Time-Dependent Effects of Urea and/or TMAO on LDH Activity 2665

Brown, A. D., and J. R. Simpson. 1972. Water relations of sugar-tolerant Santoro, M. M., Y. Liu, S. M. A. Khan, L.-X. Hou, and D. W. Bolen. 1992.
yeasts: the role of intracellular polyols. J. Gen. Microbiol. 72:589 –591. Increased thermal stability of proteins in the presence of naturally
Burg, M. B., E. D. Kwon, and E. M. Peters. 1996. Glycerophosphocholine occurring osmolytes. Biochemistry. 31:5278 –5283.
and betaine counteract the effect of urea on pyruvate kinase. Kidney Int. Somero, G. N. 1986. Protons, osmolytes, and fitness of internal milieu for
50:S1–S5. protein function. Am. J. Physiol. 251:R197–R213.
Cho, I. C., and H. Swaisgood. 1973. Factors affecting tetramer dissociation Stewart, G. R., and J. A. Lee. 1974. The role of proline accumulation in
of rabbit muscle lactate dehydrogenase and reactivity of its sulfhydryl halophytes. Planta. 120:279 –289.
groups. Biochemistry. 12:1572–1577. Wang, A., and D. W. Bolen. 1996. Effect of proline on lactate dehydro-
Corbett, R. J. T., and R. S. Roche. 1984. Use of high-speed size-exclusion genase activity: testing the generality and scope of the compatibility
chromatography for the study of protein folding and stability. Biochem- paradigm. Biophys. J. 71:2117–2122.
istry. 23:1888 –1894. Wang, A., and D. W. Bolen. 1997. A naturally occurring protective system
de Meis, L. 1988. Effects of organic solvents, methylamines, and urea on in urea-rich cells: mechanism of osmolyte protection of proteins against
the affinity for Pi of the calcium-ATPase of sarcoplasmic reticulum. urea denaturation. Biochemistry. 36:9101–9108.
J. Biol. Chem. 263:157–161. Withycombe, W. A., D. T. Plummer, and J. H. Wilkinson. 1965. Organ
Forster, R. P., and L. Goldstein. 1976. Intracellular osmoregulatory role of specificity and lactate-dehydrogenase activity: differential inhibition by
amino acids and urea in marine elasmobranchs. Am. J. Physiol. 230: urea and related compounds. Biochem. J. 94:384 –389.
925–931. Yancey, P. H. 1985. Organic osmotic effectors in cartilaginous fishes. In
Garcia-Perez, A., and M. B. Burg. 1990. Importance of organic osmolytes Transport Processes: Iono- and Osmoregulation. Springer-Verlag, Ber-
for osmoregulation by renal medullary cells. Hypertension. 16:595– 602. lin, Heidelberg. 424 – 436.
Gopal, S., and J. C. Ahluwalia. 1993. Effect of osmoregulatory solutes on Yancey, P. H., M. E. Clark, S. C. Hand, R. D. Bowlus, and G. N. Somero.
the stability of proteins. J. Chem. Soc. Faraday Trans. 89:2769 –2774. 1982. Living with water stress: evolution of osmolyte systems. Science.
Hand, S. C., and G. N. Somero. 1982. Urea and methylamine effects on 217:1214 –1222.
rabbit muscle phosphofructokinase. J. Biol. Chem. 257:734 –741. Yancey, P. H., and G. N. Somero. 1978. Urea-requiring lactate dehydro-
Laidler, K. J. 1965. Chemical Kinetics. McGraw-Hill, New York. genases of marine elasmobranch fishes. J. Comp. Physiol. 125:135–141.
Lin, T.-Y., and S. N. Timasheff. 1994. Why do some organisms use a Yancey, P. H., and G. N. Somero. 1979. Counteraction of urea destabili-
urea-methylamine mixture as osmolyte: thermodynamic compensation zation of protein structure by methylamine osmoregulatory compounds
of urea and trimethylamine N-oxide interactions with protein. Biochem- of elasmobranch fishes. Biochem. J. 183:317–323.
istry. 33:12695–12701. Yancey, P. H., and G. N. Somero. 1980. Methylamine osmoregulatory
MacMillen, R. E., and A. K. Lee. 1967. Australian desert mice: indepen- solutes of elasmogranch fishes counteract urea inhibition of enzymes.
dence of exogneous water. Science. 158:383–385. J. Exp. Zool. 212:205–213.