Proc. Nat. Acad. Sci. USA Vol. 79, pp. 7180-7184, December 1982 Biochemistry Conformational implications of enzymatic proline hydroxylation in collagen (prolyl hydroxylase/(-turn/hydroxyproline/collagen folding) R. K. CHOPRA AND V. S. ANANTHANARAYANANt Department ofBiochemistry, Memorial University of Newfoundland, St. John's, Newfoundland, Canada AlB 3X9 Communicated by Walter Kauzmann, August 30, 1982 ABSTRACT In 1979 it was proposed that prolyl hydroxylase the enzymatic hydroxylation process. The results appear to ver- (prolyl-glycyl-peptide,2-oxoglutarate:oxygen oxidoreductase, EC ify the postulates on the conformational criterion for and the 1.14.11.2) recognizes the 3-turn conformation in nascent procol- consequence ofproline hydroxylation in collagen. Additionally, lagen chains and that the hydroxylation process involves a con- they indicate an enhanced rate of folding of the hydroxylated formational change resulting in "straightening" ofthe (turn seg- polypeptide chains into the collagen-like triple-helix. Prelimi- ments into the linear triple-helical conformation ofnative collagen. nary reports of our findings have We present experimental data that verify both these postulates. appeared (4, #:). The following peptides were synthesized and studied for their con- formation and interaction with prolyl hydroxylase: tBoc-Pro-Gly- MATERIALS AND METHODS Ala-OH, tBoc-Pro-Gly-Val-OH, tBoc-Gly-Val-Pro-Gly-Val-OH, Enzyme. Prolyl hydroxylase from 13-day chicken embryos and tBoc-Pro-DAla-Ala-OH. Spectral data showed that these pep- was purified to homogeneity by using a recent modification (5) tides preferred a (-turn conformation. All of them acted as in- of the hibitors ofthe enzyme; the pentapeptide also acted as a substrate. affinity chromatography procedure (6). The enzyme ac- To mimic the biosynthetic event, a collagen model polypeptide, tivity (expressed in ,umol of hydroxyproline produced per mg (Pro-Pro-Gly)10, was incubated at 37C with purified prolyl hy- ofprotein per hr) ranged from 60 to 90 units. droxylase and the necessary cosubstrates and cofactors at pH 7.8. Peptides and Chemicals. (Pro-Pro-Gly)54H20 and (Pro- A progressive change from the initially nonhelical to the triple- Pro-Gly)5'9H20 were obtained from Proteins Research Insti- helical conformation, as monitored by CD spectra and gel filtra- tute (Osaka, Japan). tBoc-Pro-Gly-Ala-OH, tBoc-Pro-Gly-Val- tion, occurred during the course of proline hydroxylation. In ad- OH, tBoc-Pro-D-Ala-Ala-OH, and tBoc-Gly-Val-Pro-Gly-Val- dition to leading to increased thermal stability ofthe triple-helical OH were synthesized and characterized by using standard pro- conformation in (Pro-Pro-Gly)10 and (Pro-Pro-Gly)5, the enzy- cedures (7, 8). Catalase, bovine serum albumin, a-ketoglutaric matic incorporation of the hydroxyproline residues was found to acid, ferrous sulfate, dithiothreitol, ascorbic acid, and Tris-HCI enable these polypeptides to fold into this conformation faster than were obtained from Sigma. a-[1-14C]Ketoglutaric acid and scin- the unhydroxylated counterparts. These conformational implica- tillation fluids were obtained from New England Nuclear; and tions ofproline hydroxylation in collagen may also be ofuse in the Sephadex G-50 was from Pharmacia (Uppsala, Sweden). Deion- study of the complement subcomponent Clq and of acetylcholine ized water was used throughout. All other chemicals were of esterase which contain collagen-like regions in them. analytical grade. Enzyme Assay and Proline Hydroxylation. The enzyme ac- The post-translational hydroxylation of selected proline resi- tivity was determined by measuring the 14CO2 evolution due dues by prolyl hydroxylase (prolyl-glycyl-peptide,2-oxoglutar- to the stoichiometric oxidation of labeled a-ketoglutaric acid ate:oxygen oxidoreductase, EC 1.14.11.2) is a crucial event in during proline hydroxylation (6, 9). The synthetic tripeptides the biosynthesis ofcollagen. The resulting hydroxyproline res- and pentapeptides were tested for enzymatic proline hydrox- idues are essential for the stability of the triple-helical confor- ylation by replacing the (Pro-Pro-Gly)5 substrate in the proce- mation of collagen at body temperature (1). In regard to the dure for enzyme assay (6) with the respective peptide (35 mM) conformational aspects ofthe enzymatic hydroxylation process, and using 30 jig ofthe enzyme. In the inhibition studies, each two questions needed to be answered: (a) What is the preferred peptide was used at 5-10 mM and its effect on the hydroxylation conformation ofthe peptide substrate? and (b) What is the con- of (Pro-Pro-Gly)5 was studied at varying concentration of the formational consequence of proline hydroxylation? Based on latter (0.19-1.85 mM in tripeptide units). experimental data and theoretical considerations, it was pro- CD Spectral Measurements. These were made with a Jasco posed (2) that prolyl hydroxylase recognizes the (3-turn confor- J-20 spectropolarimeter in water-jacketed cells of 1-5 mm path mation (3) formed at the -Pro-Gly- segments in the nascent pro- lengths. The conformational change during the enzymatic pro- collagen chains and that the hydroxylation process results in line hydroxylation of (Pro-Pro-Gly)10 was studied as follows. In "straightening" of these segments into the rigid conformation a conical flask, 2-4 ml of a solution containing polypeptide at necessary for the subsequent association of these chains in the 0.5 mg/ml (0.95 mM in tripeptide) previously heated to 1000C triple-helical conformation of native collagen. for 10 min and chilled in ice, catalase at 0.1 mg/ml, 1.0 mM In order to test these postulates in a direct manner, we un- ascorbic acid, 0.05 mM ferrous sulfate, 0.1 mM dithiothreitol, dertook (a) the synthesis of simple peptides which were ex- and prolyl hydroxylase at 30-50 ,ug/ml in 50 mM Tris-HCl buff- pected to prefer the (3-turn conformation to study their inter- er at pH 7.8 was incubated at 370C. a-Ketoglutaric acid (2.0 actions with prolyl hydroxylase and (b) the study of the mM) containing the "'C label (200,000 dpm) was then added, conformational change in polypeptide models ofcollagen during the solution was mixed well, and a stopwatch was turned on. The publication costs ofthis article were defrayed in part by page charge t To whom correspondence should be addressed. payment. This article must therefore be hereby marked "advertise- t Chopra, R. K. & Ananthanarayanan, V. S., Twelfth International Con- ment" in accordance with 18 U. S. C. §1734 solely to indicate this fact. gress of Biochemistry, Perth, Australia, Aug. 15-21, 1982. 7180 Downloaded by guest on September 24, 2021 Biochemistry: Chopra and Ananthanarayanan Proc. Natl. Acad. Sci. USA 79 (1982) 7181 1 S FIG. 1. Lineweaver-Burk plotforthe hydroxylation of (Pro-Pro-Gly)5 (0.20-2.0 mM)by prolyl hydroxylase (1 /g) in the absence (io) and presence of the synthetic peptides tBoc-Pro-DAla-Ala-OH (A), tBoc-Gly-Val-Pro-Gly-Val-OH (m), tBoc-Pro-Gly-Val-OH (0), and tBoc-Pro-Gly-Ala-OH (i). The rate v15 represents mol of hydroxyproline formed in 15 min per mg of the enzyme. The substrate concentration is in mg/ml. An aliquot (50 ,ul) ofthe solution was quickly removed, treated 8-turn which was predicted forpeptides with the L-D sequence with dilute sulfuric acid, neutralized with alkali, and mixed with (11). This conformation was found in both nonaqueous solvents scintillation fluid for measurement of 14C activity of the un- and water. The pentapeptide tBoc-Gly-Val-Pro-Gly-Val-OH reacted a-ketoglutaric acid. Another aliquot (about 0.2 ml) was was found to be similar to the tripeptide tBoc-Pro-Gly-Val-OH injected into the empty CD cell kept at 370C, the spectrum was and existed predominantly in the type II }-turn conformation recorded (between 220 and 240 nm), and the time was noted in trifluoroethanol. when spectral recording at 226 nm was made. The cell was then The interaction ofthe synthetic peptides with purified prolyl emptied and refilled with additional aliquots from the original hydroxylase was studied by assessing their susceptibilities to solution and the spectra were recorded as above. proline hydroxylation and their capacities to inhibit the hy- Continuous monitoring of the CD spectrum of the same so- droxylation of the standard synthetic substrate (6), (Pro-Pro- lution was not feasible because the hydroxylation reaction did Gly)5. The tripeptides did not undergo any significant hydrox- not proceed well inside the narrow cell due to insufficient avail- ylation; the pentapeptide tBoc-Gly-Val-Pro-Gly-Val-OH was ability of oxygen, even when the cell was not stoppered. The hydroxylated with a Km of 25 mM (of tripeptide unit) based on extents of proline hydroxylation at the various time intervals a Lineweaver-Burk plot of the kinetic data. This pentapeptide were also estimated in a separate experiment using the 14Co2 as well as the three tripeptides acted as inhibitors of (Pro-Pro- method (9) and the direct determination of hydroxyproline (10). Gly)5 to varying extents (Fig. 1). The details ofthe kinetic stud- These were found to correlate well with the estimates obtained ies will be given elsewhere. by measuring the radioactivity of the unreacted a-ketoglutaric Conformational Change During Proline Hydroxylation. acid. Fig. 2 shows the CD spectra of the incubation mixture at various time intervals. The initial spectrum is mainly due to (Pro-Pro- RESULTS Gly)10 in the nonhelical form (resulting from the prior heating Conformation of Synthetic Peptides and Their Interaction of the polypeptide to 1000C and the subsequent chilling). [Pro- with Prolyl Hydroxylase. The conformation of the synthetic lyl hydroxylase has a considerable amount (>40%) ofthe a-he- peptides was determined from CD, IR, and NMR data in lical conformation which contributes to the CD in the experi- aqueous and nonaqueous solvents as in the earlier studies (7, mental wavelength region (unpublished data); however, this 8).