Long chain enoyl coenzyme A hydratase from pig heart.

Abstract A long chain enoyl-CoA hydratase, which in addition to crotonase is present in pig heart muscle, has been isolated and partially purified. This enzyme appears to be located in the mitochondria and is tighter membrane-bound than crotonase. It catalyzes the hydration of medium and long chain trans-Δ2,3-enoyl-CoA substrates to their corresponding l-3-hydroxy derivatives but it is nearly inactive toward crotonyl-CoA. The highest Vmax was observed with Δ2,3-octenoyl-CoA, while longer chain substrates gave progressively decreasing values. However, the same Km value of 24 µm was obtained for all even numbered Δ2,3-enoic acid derivatives containing 8 to 14 carbons. The pH optimum of this enzyme was found to be 8.5. Inhibition studies showed that this hydratase, in contrast to bovine liver crotonase, is not significantly inhibited by acetoacetyl-CoA but is strongly inhibited by long chain enoylCoA substrates. This inhibition can be at least partially prevented by the presence of bovine serum albumin. The enzyme is also inhibited by sulfhydryl inhibitors, as, for example, p-chloromercuribenzoate or N-methylmaleimide, which at concentrations of 1 mm inhibited the enzyme to the extent of 100% and 69%, respectively.


Long Chain Enoyl Coenzyme
A Hydratase from Pig Heart* (Received for publication, September 14, 1973) HOIST SCHULZ From the Departnlent of Chemistry, City College of the City C:nivemity of n'ew York, h-ew York, New York 10031 SUMMARY A long chain enoyl-CoA hydratase, which in addition to crotonase is present in pig heart muscle, has been isolated and partially purified.
This enzyme appears to be located in the mitochondria and is tighter membrane-bound than crotonase.
It catalyzes the hydration of medium and long chain trans-A2*3-enoyl-CoA substrates to their corresponding L-3-hydroxy derivatives but it is nearly inactive toward crotonyl-CoA.
The highest V,,,,, was observed with A2f3octenoyl-CoA, while longer chain substrates gave progressively decreasing values. However, the same K, value of 24 pM was obtained for all even numbered A213-enoic acid derivatives containing 8 to 14 carbons. The pH optimum of this enzyme was found to be 8.5.
Inhibition studies showed that this hydratase, in contrast to bovine liver crotonase, is not significantly inhibited by acrtoacetyl-CoA but is strongly inhibited by long chain enoyl-CoA substrates.
This inhibition can be at least partially prevented by the presence of bovine serum albumin.
The enzyme is also inhibited by sulfhydryl inhibitors, as, for example, #-chloromercuribenzoate or N-methylmaleimide, which at concentrations of 1 mM inhibited the enzyme to the extent of 100% and 69%, respectively. the assumption that crotonasc is the only hydratase involved in /3 oxidation. The present report describes the partial purification and characterization of a long chain enoyl-Co-1 hydratase from pig heart. This enzyme is virtually inactive with crotonyl-CoA, but it shows comparable activities toward different longer chain enoyl-Coh compounds.
It appears to be located in the mitochondria and is more tightly associated with membranes than is crotonase.

2705
The concentrations of A2*3-enoyl-CoA substrates were determined in several ways: (a) by use of the method of Ellman (7) after cleaving the thioester bond with hydroxylamine at pN 7; (5) by determining the decrease in absorbance at 263 nm after cleaving the thioester bond with hydroxylamine at pH 7; and (c) by determining the decrease in absorbance at 263 nm as a result of the crotonase-cata!yzed hydration of the substrates. However, the third method is of limited value because the equilibrium constants have only been reported for crotonyl-Coil and A2v3-hexenoyl-CoA (1,8,9). A molar extinction coefficient of ~263 = 6700 cm-i M-' was used for calculating the substrate concentrations.
.%zyme Assays-The hydratase activity was routinely measured by following the decrease in absorbance at 263 nm on a Gilford recording spectrophotometer, model 240 (direct method). A standard assay contained 50 pmoles of potassium phosphate (pH 8), 50 pg of bovine serum albumin, 15 nmoles of AZ+*-enoyl-CoA, and 2 to 4 pg of partially purified long chain enoyl-CoA hydratase in a total volume of 0.6 ml. Under these conditions the rates were linear for approximately 2 min. In some instances a combined assay was used in which the hydratase was coupled with the NAD+-dependent n-3-hydroxylacyl-CoA dehydrogenase. When an excess of dehydrogenase was used, the reduction of NAD+ was dependent only on the rate of hydration of the A2*3-enoyl-CoA substrate.
Purification of Long Chain Enoyl-CoA Hyclratase-Frozen pig heart (225 g) was cut into small pieces and blended together with 600 ml of 0.05 M potassium phosphate (pH 7.0) and 5 mM mercaptoethanol for 5 min. The resulting suspension was centrifuged for 30 min at 27,000 X g and the precipitate discarded.
The supernatant (homogenate) was fractionated with (NH,)&OI and the protein fraction which precipitated between 30 and 75% saturation was collected by centrifugation and dissolved in 0.01 M potassium phosphate (pH 7.0) and 5 mM mercaptoethanol.
This solution was extensively dialyzed against 0.01 M potassium phosphate (pH 7.0) and 5 mM mercaptoethanol and was then applied to a DEAEcellulose column (5 X 45 cm) which had been previously equilibrated with 0.01 M potassium phosphate (pH 7.0) and 5 mM mercaptoethanol.
The column was washed with 0.01 M potassium phosphate (pH 7.0), 5 mM mercaptoethanol, and 0.1 M NaCl until all ultraviolet-absorbing material ceased to be eluted. The column was then developed with a gradient made up of 1.6 liters each of 0.01 M potassium phosphate (pH 7.0)-5 mM mercaptoethanol-0.1 M NaCi and 0.01*~ potassium phosphate (pH 7.0)-5 mM mercantoethanol-0.6 M NaCl. Fractions of 25 ml were collected and assayed for hydratase activity with the CoA derivatives of crotonic acid, A2s3-decenoic acid and either A2*3-hexadecenoic acid or Aa*3-tetradecenoic acid. The fractions containing long chain hydratase free of apparent crotonase activity were pooled and the protein was precipitated with (NH,)&OI and finally dissolved in a minimal volume of 0.01 M potassium phosphate (pH 7.0) and 5 mM mercaptoethanol.
Samples of such partially purified hydratase were stable for several months when stored at--zoo. Subcellular Fractionation of Pia Heart Homoaenate-For subcellular fractionations small" pieies of pig heart were forced through a meat grinder, suspended in 0.25 M sucrose, and homogenized in a Teflon-pestled Potter-Elvehjem type homogenizer attached to a motor spun at 1,200 rpm. The homogenate was filtered through a glass wool plug and fractionated in a similar manner as described by Schneider and Hogeboom for rat liver homogenates (10). The following five fractions were obtained under the indicated conditions: nuclei at 500 X g for 10 min; heavy mitochondria at 5,000 X g for 20 min; light mitochondria at 24,000 X g for 10 min; microsomes at 122,000 X g for 60 min, and the 122,000 X g supernatant or soluble fraction. The enoyl-CoA hydratase activities were determined with crotonyl-CoA, A2s3decenoyl-CoA, and A2*3-hexadecenoyl-CoA as substrates as described above. Succinate-cytochrome c reductase was assayed as described by Stotz (11). L-3-Hydroxyacyl-CoA dehydrogenase was measured spectrophotometrically at 340 nm. The assay mixture contained 50 pmoles of potassium phosphate (pH 7.0), 60 nmoles of NADH, 16 nmoles of acetoacetyl-CoA, and an aliquot of the fraction to be measured in a total volume of 0.6 ml.

Intracellular Localization of Long Chain Enoyl-CoA Hydratase-
The pig heart homogenate was separated into five subcellular fractions which were assayed with respect to five enzymatic activities.
Since the 24,000 X g pellet, named light mitochondria, was virtually devoid of any of these five activities, there will not be further reference to it. As shown in Fig. 1, succinatecytochrome c reductase, which is a membrane-bound mitochondrial activity, was found in the 5,000 x g and also in the 500 X g pellet, a finding which shows that the nuclear fraction contained a significant portion of the mitochondria.
Crotonase and L-3-hydroxyacyl-CoA dehydrogenase, both of which are soluble mitochondrial enzymes, were found chiefly in the 122,000 x g supernatant, a finding which suggests that most of the mitochondria were broken. This observation is not surprising in view of the rigorous treatment required for homogenizing the tough pig heart muscle. In contrast, the A*~3-hexadecenoyl-CoA hydratase activity was mostly (72%) associated with the mitochondrial fragments present in the 5,000 x g and 500 x g fractions, while the A2*3-decenoyl-CoA activity was evenly distributed between the 122,000 x g supernatant and the mitochondria-containing fractions.
It is likely that the A2e3decenoyl-CoA hydratase activity reflected the combined actions of crotonase and the long chain enoyl-CoA hydratase.
Thus it is concluded that the long chain enoyl-CoA hydratase is located in the mitochondria and is more tightly membrane-bound than are either crotonase or L-3-hydroxyacyl-CoA dehydrogenase. Of interest is the observation that a higher percentage of crotonase than of L-3-hydroxyacyl-CoA dehydrogenase was found in the 122,000 x g supernatant.

2706
Yield" % 100 63 12.5 membrane than is crotonase or that another acetoacetyl-CoA reductase, which would have to be more tightly membranebound than n-3-hydroxyacyl-CoA dehydrogenase, is present in pig heart mitochondria.
Isolation and Partial PuriJication oj Long Chain Enoyl-CoA Hydratase-Long chain enoyl-CoA hydratase was separated from crotonase and partially purified by ammonium sulfate fractionation and chromatography on DEAE-cellulose as summarized in Table I. Assuming that all activity observed with A2s3-hexadecenoyl-CoA was solely due to the long chain hydratase, a 8.5-fold purification was achieved. Significant amounts of long chain enoyl-CoA hydratase activity appeared to remain membrane-bound as evidenced by their presence in the 27,000 x g precipitate of the crude heart extract and in the 30% ammonium sulfate precipitate.
Also, upon dialysis of the 30 to 750/, ammonium sulfate fraction, a precipitate formed which contained substantial amounts of long chain enoyl-CoA hydratase.
These observations, in addition to the result from the subcellular fractionation experiment, strongly suggest that the long chain enoyl-CoA hydratase is associated with the mitochondrial membrane and is only partially solubilized upon homogenization in a blender.
The separation of long chain enoyl-CoA hydratase and crotonase achieved by DEAE-cellulose chromatography is shown in Fig. 2. The long chain hydratase (Peak II) was elutcd with approximately 0.45 M NaCl, and the resulting material was essentially free of crotonyl-CoA hydratase, thiolase, and 3-hydrosyacyl-CoA dehydrogenase activities.
Peak I represents crotonase but possibly also long chain enoyl-CoA hydratase which has a tendency to remain associated w&h other proteins.
For example, when during the isolation of this hydratase, the ammonium sulfate fractionation was not performed, the crotonase-free long chain hydratase was eluted at the position of the second large protein peak. The material present in the fractions corresponding to Peak II was isolated and chromatographcd on Sephadex G-200. Since the hydratase was eluted as a sharp peak directly with the void volume (data not shown), it can be concluded that either long chain enoyl-CoA hydratase is a very large protein or that it tends to form aggregates. Evidence for the latter possibility was obt.ained from the observation that a fraction of the partially purified long chain enoyl-CoA hydratase precipitated when centrifuged for 30 min at 90,000 x g. Only 66% of the hydratase remained in solution upon centrifugation but treatment Assays were performed with 2+1 aliquots from Fractions 1 to 60 and with lo-p1 aliquots from all subsequent fractions.
of the precipitate with phosphate buffer containing 1% Triton X-100 resulted in solubilization of 22yo of the original activity. The percentage of the precipitated hydratase increased when the enzyme was kept for prolonged periods in dilute solution. For this reason and because of an observed increase in substrate inhibit.ion with increasing purity, the enzyme was not further purified.
Catalytic Properties of Long Chain Enoyl-CoA Hydratase-The long chain enoyl-CoA hydratase catalyzes the hydration of A2~3-cnoyl-CoA substrates to their corresponding L-3-hydroxy derivatives as evidenced by the ability of the hydration products to serve as substrat.es for n-3-hydroxyacyl-CoA dehydrogenase (see Table II). The reduction of 1 mole of NAD+ per mole of A*J-decenoyl-CoA was observed when taking the 85% completion of the dehydrogenation at pH 9 (12) into account. The direct measurement of the hydration of A2~3-decenoyl-CoA at 263 nm showed that at equilibrium the ratio of 3-hydroxydecenoyl-Cob to A2v3-decenoyl-CoA was 2.8 (see Table II) from which an equilibrium constant of K = 5.1 X lop2 M was calculated.
This value is similar to that found for the hydration of crotonyl-CoA by crotonase (1, 9). Fig. 3 shows the relationship between the rate of hydration

II Product determination
The hydration of A2v3-decenoyl-CoA was followed spectrophotometrically at 263 nm as described under "Experimental Procedures," except that 50 pmoles each of Tris. HCl (pH 9) and KC1 were substituted for the normally used phosphate buffer. The reduction of NADf was measured by coupling the hydratase with the L-3-hydroxyacyl-CoA dehydrogenase assay as described under "Experimental Procedures." Assays were performed by the direct method as described under "Experimental Procedures".
Of interest is the substrate inhibition which became effective at even lower substrate concentrations when no bovine serum albumin was present in the reaction mixture.
In Fig. 4 the initial velocities versus concentrations were plotted on reciprocal coordinates for several A2J-enoyl-CoA substrates.
In contrast to crotonase, this enzyme showed similar activities toward all substrates tested except for crotonyl-Cob (not shown) toward which it was nearly inactive.
The activity with crotonyl-CoA as substrate was 1% of the activity found with A2s3-decenoyl-CoA, Since crotonasc and long chain enoyl-CoA hydratase were well separated on DEAE-cellulose (see Fig. 2)) it is not very likely that the residual activity with crotonyl-Coil was due to a contamination by crotonase.
The kinetic parameters for several A2*3-enoyl-CoA substrates of different chain lengths are listed in Table III. All K, values were the same (24 PM) with the exception of the K, for A203-hexenoyl-CoA which was nearly twice as large (45 PM) as all others. The highest maximal velocity was observed with A2J-octenoyl-CoA as the substrate while longer chain substrates gave progessively decreasing values. It appears possible that the decrease in activity with increasing chain length of the substrate was due to the above mentioned substrate inhibition which becomes more pronounced with longer chain substrates. The activity of this enzyme is strongly pa-dependent as shown in Fig. 5. The optimal pH was found to be approximately 8.5, a value which is similar to values of 9 and 9.4 reported for crotonase (1,8). With decreasing pH the activity decreases and the shape of this part of the pH curve resembles a titration curve with a midpoint at pH 6.5. A slight but definite stimulation of the hydratase activity was observed when the phosphate buffer was replaced by Tris-HCl as shown in Fig. 5.
Inhibition Studies-The sensitivity of the long chain enoyl-CoA hydratase toward substrates, especially longer chain ones, has already been mentioned.
As shown in Fig. 68, Curve 1, A2J-hexadecenoyl-CoA at a concentration of 30 pM totally inhibited the enzyme. The initial change of absorbance appears to be due mostly to a nonspecific interaction of the substrate with the protein because a similar change in absorbance was observed when only the substrate and bovine serum albumin were present in the assay mixture.
Increasing amounts of bovine serum albumin resulted in increased protection against substrate inhibition (see Fig. 6A, Curves d and 3). For comparison an identical experiment was performed with A*v3hexenoyl-CoA as substrate.
As shown in Fig. 6B even this short chain substrate inhibited the enzyme, although less pronouncedly than the long chain derivative, and again bovine serum albumin prevented the inhibition.
Since the substrate 2708 n FIG. 5. The rate of hydration of Aa*3-decenoyl-CoA as a function of pH. Assays were performed by the direct method as described under "Experimental Procedures" except that for each assay 100 pmoles of the following buffers were used: potassium phosphatecitrate (Cl), potassium phosphate (O), and TriseHCl (A).

KC1
was added to maintain a constant ionic strength. 6. The effect of bovine serum albumin on the rate of hydration of A2s3-hexadecenoyl-CoA (A) and A2*3-hexenoyl-CoA (B). Each reaction mixture contained, in a total volume of 0.6 ml, 50 pmoles of potassium phosphate (pH 8), 18 nmoles of A2*3-enoyl-CoA, 4.6 rg of long chain enoyl-CoA hydratase, and the following amounts of bovine serum albumin: 1,O fig; 2,10 pg; S,25 pg. inhibition became more pronounced the purer the hydratase, it is doubtful that this inhibition has any physiological significance. The recent observation that bovine liver crotonase is strongly inhibited by acetoacetyl-CoA (2), prompted a similar investigation with the long chain enoyl-CoA hydratase.
However, in contrast to bovine liver crotonase, acetyl-CoA at the same concentration as acetoacetyl-CoA caused a similar inhibition (17%). It is possible that the observed inhibition was of the same nature as the above mentioned inhibition of long chain hydratase by CoA derivatives of A2s3-enoic acids. Thus it is suggested that the small inhibition of this enzyme by acetyl-CoA and acetoacetyl-CoA is of no physiological significance.
Finally, the effect of sulfhydryl inhibitors on the activity of the long chain enoyl-CoA hydratase was investigated.
A 69% (85%) inhibition was observed when the hydratase was preincubated for 15 min at pH 7 in the presence of 1 mM (5 mM) N-methylmaleimide and then assayed with A2*3-octenoyl-CoA as the substrate. An identical experiment performed with p-chloromercuribenzoate as the inhibitor showed an increase in the inhibition from 13% to 100% when the inhibitor concen: tration was raised from 0.1 to 1 InM. A 16% inhibition by or-iodoacetamide at 5 mM concentration was observed, but because this experiment was performed at pH 7 the observed degree of inhibition may not be optimal.
Thus it is concluded that at least one thiol group is essential for the full activity of long chain enoyl-CoA hydratase, although these findings cannot be taken as evidence for the participation of a sulfhydryl group in the catalytic event.

DISCUSSION
In discussions of fatty acid oxidation it has always been assumed that the hydration of all A2*3-enoyl-CoA intermediates is catalyzed by only one enzyme, namely crotonase (13). However, some evidence has recently been obtained which would indicate that at least in some p oxidation systems more than one hydratase must be present.
Waterson et al. found that the purified crotonase from Clostridium acetobutylicum catalyzed only the hydrat.ion of crotonyl-CoA and A2*3-hexenoyl-CoA but that crude extracts from the same organism were active with Cq t.o Cl6 derivatives, a finding which led them to suggest that in this organism a long chain enoyl-CoA hydratase may be present in addition to crotonase (14). Wit-Peters et al. have reported that during the purification of crotonase from calf liver, the ratio of hydratase activities determined with crotonyl-CoA and A2J-hexadecenoyl-CoA increased from 5.3 to 420 (15). Thus they suggested that in addition to crotonase a long chain enoyl-CoA hydratase must be present in the homogenate.
However, Waterson and Hill found that the ratio of hydratase activities obtained with crotonyl-CoA, A2J-decenoyl-CoA, and A2vs-hexadecenoyl-CoA remained constant during the purification of crotonase from bovine liver (2). Although the presence of a separate long chain enoyl-CoA hydratase in bovine liver remains in question, the present report clearly shows that a separate long chain enoyl-CoA hydratase exists in pig heart. Even though the functional role of this long chain enoyl-CoA hydratase has not yet been established, it appears reasonable to assume that it participates in the 0 oxidation of fatty acids, especially since this enzyme is localized in the mitochondria.
Assuming that both crotonase and long chain enoyl-CoA hydratase are involved in fatty acid oxidation, a comparison of their chain lengths' specificities is of interest.
Since the pig heart crotonase has not yet been purified and characterized, the data obtained with the bovine liver enzyme are used for comparative purposes. These data show that crotonase is highly active with crotonyl-Cob but acts much more slowly on longer chain substrates.
The difference between the rates of hydration with crotonyl-Cob and longer chain enoyl-CoA as substrates is most pronounced at low substrate concentrations because the K, for crotonyl-Cob is 10 times smaller than that for A2s3-hexenoyl-CoA, the second best substrate (2). In contrast, the long chain enoyl-CoA hydratase is nearly inactive with crotonyl-CoA but is fully active with all longer chain enoyl-CoA substrates.
Hence it appears that the optimal rate of hydration of all A2*3-enoyl-CoA intermediates in fatty acid oxidation would be achieved if crotonase and the long chain enoyl-CoA hydratase act in a concerted manner.
A similar situation exist.s with respect to the mitochondrial thiolytic activity which is due to a short chain specific thiolase only act.ive on acetoacetyl-CoA (16) and a general purpose enzyme which is more active on long-than on short chain substrates (17). Since a short and long chainspecific acyl-CoA dehydrogenase is also expected to be present in pig heart (18), it is possible that the enzymes of p oxidation in heart tissue are arranged in two groups, one of which is concerned with the degradation of long chain fatty acids to medium or short chain fatty acyl-Co.4 intermediates which arc then further degraded by the group of short chainspecific enzymes. This is certainly a highly speculative model which would additionally require the presence of a scco~ld ~-3. hydrosyacyl-Co-1 dehydrogenase in heart tissue. The reason for considering such an arrangement is based 011 the assumption that the enzymes of fatty acid oxidation exist in the mitochondria in a highly ordered manner and 011 the observation that t.he long chain enoyl-CoA hydratase is much more tightly membrane-bound than is crotonase.
The effect of acetoacetyl-CoX 011 the rate of hydration of A*Uecenoyl-CoA by long chain enoyl-('oh hydratase was studied because it has been reported that bovine liver crotonasc is inhibited by acctoacetyl-CoX (2). On the basis of the observed inhibition and substrate specificity, it has been suggested that crotonasc functions as a regulatory enzyme in fatty acid osidation (2). According to this proposal, acetoacetyl-Coh would inhibit the hydration of A 2~3-l~esatlccc~~~o~l-Coh or A2s3tetradecenoyl-Coh but not the hydration of shorter chain substrates to such an extent that these steps would become ratelimiting.
Experiments performed with whole mitochondria, especially beef heart mitochondria, appear to lend support to this proposal because the oxidation of palmitic acid but not of octanoic or butyric acid was reduced by acetoacetyl-CoA. Inhibition experiments performed with the long chain enoyl-Coh hydratase from pig heart showed that this enzyme was only slightly and unspecifically inhibited by acctoacctyl-CoA. Thus it rqpears highly unlikely that the proposed regulatory mechanism is effective in pig heart, if OIIC assumes that the long chain hydratase is involved in /3 oxidation. ~111 additional reason for a reevaluation of the proposed regulatory schema of fatty acid oxidation is the observed strong inhibition of L-3hydrosyacyl-Co.4 dchydrogenase by acetoacctyl-Coh with a KI of 7.7 par,' while the KI for the inhibition of crotonasc by acctoacetyl-Cob is 30 phf (2). The long chain enoyl-Co8 hydratasc was found to be inhibited by sulfhydryl inhibitors.
Although this finding does not constitute proof for the participation of a sulfhydryl group in the catalytic event, it suggests that at least one thiol group 1 J. Schifferdecker and H. Schulz, unpublished result. is positioned near or at the catalytic site or that a sulfhydryl group is required for maint,aining the enzyme in an active conformation.
Studies by 20) with crotonase have led to the suggestion that this enzyme, despite its inactivation by thiol inhibitors, does not possess a sulfhydryl group which participates in the catalytic event, but that the chemical modification of one specific thiol group results in a sterical restriction at the catalytic site which is the cause of the inhibition.
It is possible that the long chain enoyl-CoA hydratase is affected by the sulfhydryl inhibitors in a similar manner.