A conformational study of porcine thyrocalcitonin.

Abstract The structure of porcine thyrocalcitonin has been evaluated by circular dichroism, optical rotatory dispersion, infrared spectroscopy, and fluorescence. The degree of helical structure was estimated by circular dichroism (222 mµ) and optical rotatory dispersion (233 mµ) at the n → π* transition of the α helix. The optical activity at this transition suggests that thyrocalcitonin contains approximately 10% α-helical structure in aqueous solution. The spectrum of thyrocalcitonin in 6 m guanidine was similar to that of a randomly coiled polypeptide. In 2-chloroethanol, thyrocalcitonin formed a structure containing approximately 50% α helix. The near ultraviolet circular dichroic spectrum of thyrocalcitonin revealed a major band at 288 mµ, indicating that the tryptophanyl chromophore had restricted rotational freedom. Reduction and alkylation of the amino-terminal heptapeptide ring of thyrocalcitonin produced no significant change in the circular dichroic or optical rotatory spectra. In addition, the effects of alkaline pH and temperature on tryptophanyl emission were similar to the unmodified hormone. These findings preclude a high degree of organized structure in the heptapeptide ring, which is in marked contrast to the amino-terminal hexapeptide ring of oxytocin. These studies indicate that porcine thyrocalcitonin exists predominately in a random coil in aqueous solution. The polypeptide, however, does not appear to possess a rigid conformation and a coil ⇌ helix equilibrium may exist in water with guanidine shifting the equilibrium in favor of the coil and 2-chloroethanol in favor of the helix. The latter type of transition may occur at the receptor site of the hormone in lipid layers or cellular membranes where the water concentration is significantly reduced.


A Conformational
Study of Porcine Thyrocalcitonin (Received for publication, November 5, 1969) H. BRYAN BHEWER, JR., AND HAROLD EDELHOCH From the Molecular Disease Branch, National Heart Institute, and the National Institute of Arthritis and Metabolic Diseases, National Institutes of Health, Bethesda, Maryland 20014 SUMMARY The structure of porcine thyrocalcitonin has been evaluated by circular dichroism, optical rotatory dispersion, infrared spectroscopy, and fluorescence.
The degree of helical structure was estimated by circular dichroism (222 mp) and optical rotatory dispersion (233 mp) at the n + r* transition of the Q: helix.
The optical activity at this transition suggests that thyrocalcitonin contains approximately 10% a-helical structure in aqueous solution.
The spectrum of thyrocalcitonin in 6 M guanidine was similar to that of a randomly coiled polypeptide. In Z-chloroethanol, thyrocalcitonin formed a structure containing approximately 50% (Y helix. The near ultraviolet circular dichroic spectrum of thyrocalcitonin revealed a major band at 288 rnp, indicating that the tryptophanyl chromophore had restricted rotational freedom. Reduction and alkylation of the amino-terminal heptapeptide ring of thyrocalcitonin produced no significant change in the circular dichroic or optical rotatory spectra.
In addition, the effects of alkaline pH and temperature on tryptophanyl emission were similar to the unmodified hormone. These findings preclude a high degree of organized structure in the heptapeptide ring, which is in marked contrast to the aminoterminal hexapeptide ring of oxytocin. These studies indicate that porcine thyrocalcitonin exists predominately in a random coil in aqueous solution. The polypeptide, however, does not appear to possess a rigid conformation and a coil ti helix equilibrium may exist in water with guanidine shifting the equilibrium in favor of the coil and 2-chloroethanol in favor of the helix. The latter type of transition may occur at the receptor site of the hormone in lipid layers or cellular membranes where the water concentration is significantly reduced.
Porcine thyrocalcitonin is a single chain 32 amino acid polypeptide hormone of known amino acid sequence (l-3).
The peptide contains a l-7 amino-terminal disulfide bridge, carbosylterminal residue of prolinamide, and a single tyrosine and tryptophan residue, which are adjacent in the amino acid sequence. The solubility of the hormone is small since there are relatively few charged residues.
Initial studies relat,ing chemical structure to biological function have revealed that chemical modification of the single cystine, tryptophan, or t,yrosine residue results in almost complete loss of biological, but in little loss of immunological activity (4,5). Modification of the single methionine residue, however, produces no significant loss of either biological or immunochemical reactivity (4,5). Previous studies have also indicated that purified TCY may lose biological activity at alkaline pH (6), and with storage at 0" (7).
The purpose of this communication is to evaluate the secondary and tertiary structure of porcine thyrocalcitonin, and the role that its conformation may play in the biological and immunological activity of the hormone.

EXPERIMENTAL PROCEDURE
Porcine Thyrocalcitonin Preparation-Porcine TC was isolated in a homogeneous form as previously reported (8). Biological activity was 200 MRC units2 per mg as measured by rat bioassay (9). Protein concentrations were determined by absorbance measurements at 280 rnp 0.1 M acetic acid (E:, = 7570, E 1 mg/ml 1CIll = 2.1) or by amino acid analysis. Reduction of TC followed by alkylation with iodoacetic acid was performed as outlined previously (8).
Circular Dichroism-The Cary model 60 spectropolarimeter equipped with a Pockels cell was used. The optical density of all solutions was close to 2 at 280 mp. Determinations were made in l-cm cells in the near ultraviolet region, and with 0.5to 2.mm cells in the far ultraviolet region. A spectral curve was obtained on each solvent either before or after each sample. The temperature of solutions was maintained at 25.0" unless otherwise stated.
A thermistor probe was used to monitor the temperature of the cell in studies involving the temperature dependence of ellipticity.
When the effect of temperature was studied, measurements were made 30 min aft,er the equilibrium temperature had been obtained. The reduced mean residue ellipticity was obtained from the following equation: where 0 is in degrees, I in centimeters, and c in moles of residue per liter. Measurements were corrected for refractive index at n:', but not for dispersion.
Optical Rotatory Dispersion--The Cary model 60 spectropolarimeter was used. The ol)tical density of all solutions was close to 2 at 280 rnp, and determinations were made in 0.5-to 2-mm Issue of May 10, 1970 H. B. Brewer, Jr., andH. Edelhoch 2403' cells. The temperature of all solutions was 25.0". The reduced mean residue rotation was obtained from the following equation: where 01 is in degrees, I iu centimeters, and c in moles of residue per liter. Measurements were corrected for refractive index at nz", but not for dispersion.
Infrared Spectroscopy-A Beckman model IR 7 infrared spectrometer calibrated against water was used. Determinations were made on 0.1 mg of TC in a potassium bromide pellet.
Fluorescence IMeasurements-Emission spectra were obtained with a Turner model 210 recording spectrofluorometer (lo), with l-cm2 cells. Solutions had optical densities of less than 0.02 at the wave length of excitation (278 mp). The sample compartment, was equipped with a water jacket, and was maintained at 25" mlless otherwise indicated.
In experiments where the tcmperature was varied, the fluorescence measurements were determined 15 to 30 min after the equilibrium temperature had been reached. Fluorimetric p1-I curves were obtained by titrating TC with small amounts of 0.5 hf KOH or HCl from an agla syringe, while the solution was stirred magnetically.
pH determinations were made on a Radiometer model 25 pH meter. Excitation was at 278 rnp, and emission was observed at 350 m,u. The absorbance of TC solutions remains constant with pH since 278 rnp is an isosbestic point of tyrosyl ionization.
Polarization of Fluorescence-Fluorescence polarization was determined in an Aminco-Bowman spectrofluorometer as described by . Polarization, P, is defined as (Z, -GZx)/(ZV + GZx) where IV and ZH represent the intensity of vertically and horizontally polarized emission, exciting with vertically polarized light.
G is a grating correction factor defined as ZV/ZE utilizing horizontally polarized excitating light. Excitation was at 270 rnp, and emission intensities were measured at 350 mp. Polarization of samples was determined in 95% glycerol as a function of temperature (lo-40"), and as a function of glycerol concentration (75 to 95%) at 25" (12). In the experiments where the temperat,ure was varied, the fluorescence measurements were made 15 to 30 min after the temperature of the solution became constant.
The Pcrrin equation for rigid spherical molecules is as follows: where PO is the polarization at T/q ---f 0; 7, the viscosity in poise; RT, the thermal energy; 7, the excited lifetime of the molecule; and V the molecular volume.
The relaxation time (ph) was determined from the slope and intercept of a plot of ((l/Z') -(l/3)) cersus (TT/~]), in accord with the Perrin equation:  product had little absorbance in the 200 to 260 rnp range. Glycerol was spectroquality, and was obtained from Coleman and Bell. Other compounds were reagent grade, and glass distilled water was used throughout.

Circular Dichroisrn
Far Ultraviolet-The CD spectrum of native thyrocalcitonin in the wave length region of 200 to 260 rnp is reproduced in Fig. 1. The major dichroic activity is centered near 200 rnp ([0'12,, = -9,700) the region of the ellipticity band of the peptide bond in an unorganized polypeptide (14). There is also a second, weaker dichroic band near 222 rnF ([e'],, = -3,300), the wave length of the n --f r* transition of the Q! helix, which forms a shoulder on the stronger band. The CD spectrum of reduced and alkylated TC was measured in order to determine whether the disulfide group of the heptapeptide ring contributed to the spectrum of the native hormone.
The spectra of native and reduced and alkylated TC were indistinguishable ( Fig. 1). Since the minor band at 222 rnp was unaffected by reduction of the disulfide bond, its origin was investigated by measuring the CD spectrum in 6 w guanidine, a solvent iu which little or no organized structure remains in proteins (15). In 6 M guanidine the shoulder disappeared and the ellipticit,y at 222 rnp decreased to -1,100, while the major band appears to maintain its rotational strength ( Fig. 1). In order to see if the TC molecule is capable of forming a helical structure, CD measurements were performed in 2-chloroethanol, an (Y helix inducing solvent. A completely new spectrum was observed ( Fig. 1)  ],~ = --K&500), which correspond to electronic transitions associated with the peptide bond in c-r-helical polypeptides (14). Near Ultraviolet-In Fig. 2 is shown the CD spectra of TC in t,he near ultraviolet region.
The spectrum of native TC in aqueous solution, pH 3.1, has a major positive peak at 288 rnp ([VI,, = 1800) and a minor one at 280 rnp which is about half as strong.
There was almost no dichroic activity at 275 rnp, the region of tyrosyl activity.
There was negligible optical activity between 240 and 255 mM. Reduced and alkylated TC gave a CD spectrum ([WI,,, = 1600) similar to that of the unmodified hormone (Fig. 2). The CD spectrum of TC in 6 M guanidine was similar to that in water, although there was an npproximately 50% reduction in the ellipticities due to the tryytophanyl and phenylalanyl residues. The dependence of the tryptophanyl CD spectrum on ternperature is illustrated in Fig. 3. There is a progressive decrease in ellipticity at 288 rnp as the temperature is increased from 30-45".
Further increase in temperature produced only trivial the n -+ 7r* transition of the (Y helix (17), was -2600.
No significant change was found in the ORD of the reduced and alkylated hormone (Fig. 4). The ORI) spectrum of TC in 6 M guanidine (Fig. 4) resembled that of a random chain polypeptide (17). Its optical rotation at 233 rnp was -1800, the cross over disappeared and the major negative band now occurred below 210 mp. The spectrum of TC in 2-chloroethanol is also included in Fig. 4. An appreciable increase was found in the major band centered near 233 rnl.c ([m'] 233 = -7500), and the cross over was at 222 rnfi.

Infrared Spectroscopy
The infrared spectrum of native thyrocalcitonin in the region of the amide I band is illustrated in Fig. 5. The amide I band is near 1658 cm-l, the frequency associated with polypeptides predominately in unordered conformations (18). No significant bands were present at 1632 and 1685 cm-l, the frequencies characteristic of /3 structure (pleated sheet).

Emission Spectra
The emission spectrum of native TC at pH 3.1 revealed a maximum at 348 rnp (Fig. 6), which is characteristic of the indole spectrum of tryptophan or simple tryptophanyl peptides in aqueous solution (19). There was no significant fluorescence at 300 mp, the region of tyrosyl emission.
This result was not unexpected since the tyrosyl (Residue 12) and tryptophanyl (Residue 13) chromophores are neighbors and energy transfer can occur from the former to the latter with a high efficiency as has been found in the dipeptide, tryptophanyl tyrosine (20). There were only minor, insignificant changes in the emission spectrum or quantum yield of TC in 6 M guanadine when compared with the data in water (Fig. 6).

Effect of pH
A&d-The fluorescence intensity of native TC, measured at 348 rnp, decreased uniformly by about 25% when acidic solutions of TC were neutralized to pH 7.5 (Fig. 7). It is uncertain whether this decrease represents an intramolecular change or is the result of molecular association even though a similar titration curve was obtained in a solution containing twice the concentration of TC. When solutions were back-titrated from neutrality to pH 2.5 the fluorescence intensity returned to its initial value. The loss in fluorescence with neutralization suggests that a pHdependent modification in the behavior of the tryptophanyl reeidue occurs since titration of other polypeptide hormones, which appear to be predominately random chains, revealed a small increase in fluorescent intensity in the same pH region (21). The pa-dependent decrease in fluorescence disappeared in reduced and alkylated TC.
In solutions of TC containing 3 M urea or 6 M guanidine there were only trivial changes in emission intensity between PI-I 2.5 and 7.5 (Fig. 7). These results may reflect either a loss of weak intrachain tryptophanyl interactions in the molecule or an increase in TC solubility.
)lIkali-The loss of tryptophanyl emission by radiationless energy transfer to ionized tyrosine has been studied in synthetic peptides, polypeptide hormones, and in proteins (20-23).
The degree of quenching was found to decrease with the distance between the chromophores in a homologous series of pept,ides of tlyptophan and tyrosine (20). The quenching of tryptophanyl emission with tyrosyl ionization in TC is illustrated in Fig. 8. The high degree of alkaline quenching is in accord with the evidence of efficient transfer from tyrosine to tryptophan suggested by the complete absence of tyrosyl emission at acid pH in both water and guanidine solutions (Fig. 6). r\To significant difference was observed in the pH-fluorescence behavior in alkali between solutions of native TC in water and 6 M guanidine or aqueous solutions of reduced and alkylated TC.
The pK of the tyrosyl residue determined fluorometrically by quenching of tryptophanyl fluorescence was near 10 (Fig. 8), which is in agreement with the pK of the tyrosyl group determined by spectrophotometric titration at 245 rnp (8). The normal value of phenolic ionization precludes strong illteract,ions of this residue with other residues in the polypeptide chain.

Effect of Temperature
The thermal dependence of fluorescence emission has been studied in tryptophan and tyrosine, synthetic polypeptides of tryptophan and t,yrosine, polypeptide hormones, and proteins (21,22,24). Any deviation from a monotonic decline in emission intensity with increasing temperature implies a structural alteration. The temperature-fluorescence profile of tryptophanyl emission of TC at pH 3.1 is shown in Fig. 9. A monotonic decline was observed over the temperature range of 25-65". Reduced and alkylated TC showed a similar decline in emission intensity with increasing temperature (Fig. 9).
Polarization 0,f Fluorescence The variation of polarization of native and reduced and alkylated TC as a function of glycerol concentration is shown in Fig. 10. The data for the 2 molecules are indistinguishable, indicating that cleavage of the heptapeptide ring has no effect on the rotational properties of the indole chromophore. The lifetimes of the native hormone in 95, 85, and 75:; glycerol rvere 4.6, 4.1, and 3.6 X lo+' set, respectively.
This value is larger than that suggested as the maximum for the indole group in a structureless polppeptide chain 02).
Measurements of the variation of polarization of TC in 95'g (w/w) glycerol with temperature gave an intercept similar to the one reported above although the relaxation time was slightly smaller.

DISCUSSION
Secondary Structure--It is of interest in understanding the biological and chemical behavior of porcine TC to know if it possesses organized structure.
It is not immediately apparent whether a molecule of 32 amino acid residues will form an ordered structure in solution.
Measurements of ORD, CD, infrared spectroscopy, and fluorescence can indicate whether the polypeptide possesses hydrogen-bonded or other interactions. These parameters, therefore, have been utilized in order to gain some appreciation of the extent of organization in porcine TC.
The degree of secondary structure in TC was assessed by CD, ORD, and infrared spectroscopy.
The principal CD band of TC in the far ultraviolet is centered near 200 rnp which is the wave length corresponding to the g ---f r* transition of the random form of the peptide bond. A weaker dichroic band is present at 222 nip, the region of the n + 71-* transition of the peptide bond in its a-helical form. This degree of ellipticity, if caused by helical structure, would indicate that TC contains approximately 10% o( helix if one accepts a [0'] value of -30,400 for this transition in an (I! helix (14). It should be pointed out that t,he aromatic chromophores also have transitions in this wave length region and may also contribute to the ellipticity (25). The CD spectrum of TC is normalized in 6 M guanidine and approaches that of unordered polypeptides.
Similar results were obtained when the degree of helical structure was evaluat'ed by ORD.
The mean molar residue rotation at 233 rnp was by guest on March 24, 2020 http://www.jbc.org/ Downloaded from 240'7 -2,400; if caused by a-helical structure, this would indicate that 'K contains about 10% helix assuming a [m'] value of -13,000 and -1,800 for the transit,ion between an 01 helix and random chain (17). The band at 233 rnp was essentially eliminated in 6 M guanidine and there ~vas a spectral shift at lower wave lengths from a positive to :I negative band. The spectrum in 6 M guanidine is characteristic of the peptide chromophore when in a randomly coiled polypeptide (17).
The TC polypeptide chain is capable of forming a high degree of oc-helical structure as indicated by the CD and ORD spectra in 2-chloroethanol.
The ellipticity at 222 rnp, and the optical rotation at 233 rnp indicated that the molecule contained about 507; a! helix based on values of [W] and [m'] of synthetic polypeptides in their helical and random forms.
These combined results suggest that, TC in acid solution contains a very small amount of OL helix, which may correspond t,o about, one turn for a polypeptidc of 32 residues.
Tertiary Structure-The tertiary interactions of TC were also evaluated by t,he behavior of its chromophoric residues. The CD spectrum in the near ultraviolet region was determined since this region is free of contributions from the peptide chromophore which absorbs in the same ultraviolet region as the second electronic t,ransition in the aromatic groups.
The positive CD spectrum, at 288 mp, is clearly caused only by the t,ryptophanyl residue in TC. The rotational strength of the band implies that the indole residue lacks some degree of rotational freedom with respect to the asymmetric carbon of the tryptot)hanyl or to some other neighboring asymmetric cent,er. An illdole hcptapeptide ring interaction can be escluded since the magnitude of the molar ellipticit,y was not significantly affected by reduction of the disulfide bond.
The dichroic activity in 6 M guanidine and in water at 45" approaches values in linear dipeptides and may represent, its limiting value in this polypeptide (16). The positive CD spectrum from 255 to 265 nip can be assigned to the phenylalaninc residue or residues in TC (phenylalaninc residues are present at posit,ions 19, 22, and 27 in the amino acid sequence (l-3)).
The presence of a phenylalanyl dichroic band, which is significantly reduced in 6 &I guanidine, indicates that one or more of these residues lacks some rotational freedom.
The fluorescent propertics of tyrosine and tryptophan have recently been employed as an indicator of tertiary structure in polypeptide hormones (21). In TC the wave length maximum of the indole fluorescence was 348 nip, which is very close to that of model tryptophanyl peptides of low molecular weight and indicates that the fluorochromc is essentially in an aqueous environment (19). The indifference of the emission spectrum to 6 M guanidine and the monotonic decline in tryptophanyl emission with increasing temperature are in accord with this point of view. Consequently, it is unlikely that the indole residue is part of a cluster of groups stabilized by hydrophobic interactions.
The only fluorimetric evidence suggesting a perturbat,ion of the indole group in TC is the acid titration curve in water. The c:uenching of indole emission with neutralization is the reverse of the changes observed in other polypeptide hormones and in several stable proteins (21). Moreover, the pH dependence of the fluorescence of TC between pH 2.5 and 7.5 was normalized in 6 M guanidine.
The change in fluorescence may originate from t,he rupture of a weak interaction between the indole group and the side chain of a neighboring residue. It should be emphasized that the ('1) data are much more convincing than the fluorescence dat,a in showing an important interaction of the indole moiety with other parts of the hormone. This result, is not unusual since the two types of data represent different, electronic processes with minor differences in energies and major differences in lifetimes.
The important role played by the electronic configuration in the interaction of the molecule with the solvent has been reported recently in the case of two aromatic diketopiperazines.
Nuclear maglletic resonance spectra indicate that the dikctopiperazincs Gly-L-Tyr and Gly-L-'l'rp in dimethyl sulfoxide have a folded configuration in their ground state whereas they have unfolded forms in their cxcit,ed states, as revealed by fluorescence measurements (26).
Energy transfer between the indole and phenol chromot)hores has been proposed to explain the lack of tyrosyl fluorescence in TC and the quenching of tryptophanyl fluorescence in alkaline solution.
The neighboring positions of the 2 aromatic residues can readily account for the failure to observe tyrosyl fluorescence since it was not observed in Trp-Tyr whereas emission was observed in 'l'rp-Gly,Tyr which becomes stronger in Trp-Gly, Tyr (20). The quenching of tryptophanyl fluorescence by phenolate by energy transfer is a more efficient process and has been reported in numerous polypeptides and proteins (21-23).
The large value of the intercept in the I'errin plot of the polarization data presumably arises from energy transfer between illdole groups.
TC should therefore be associat,ed in glycerol solutions since it contains only a single tryptophanyl residue. The relaxation time calculated for TC is considerably greater than expected for a structureless molecule of 32 residues. A value which is half as large has been report,ed for adrenocorticotropic hormone which contains 39 residues (12). It is not clear, however, whether the large value represents organized structure arising from the association of TC molecules or whether it exists in its free state. It should also be noted that TC could acquire structure iu the concentrated glycerol solutions used in the polarization experiments, as was observed in solutions of TC in 2.chloroethanol.
There was no significant, change in either the CD or ORD spectral patterns, or in the temperature-fluorescence profile with reduced and alkylated TC when compared to the native hormone.
These results suggest that the disulfide bond does not contribute measurably to the optical activity of 'IX, and that there is little, if any, organized structure in the heptapeptide ring. Oxytocin has a similar amino-terminal disulfide bridge which is composed of 6 amino acid residues.
CD measurement on oxytocin and various analogues have revealed significant optical activity in the ultraviolet region which have been attributed to the disulfide chromophore (27,28). In oxytocin there is a marked change in ellipticity with reduction of the six-member disulfide ring. The failure to observe any dichroic activity in TC in the region of the disulfide bond transition in both the 210 and 250 rnp regions (27,28) also suggests that the heptapeptide ring does not interact with other residues of the hormone.
The marked difference in CD activity between TC and oxytocin should reflect differences in rigidity bet'ween t,he two rings. The smaller ring size or stronger residue interactions in oxytocin may account for its greater optical activity.
In summary, it appears that porcine TC exists predominately in an unordered conformation in ayueous solutions. The lack of a high degree of ordered structure in porcine TC is similar to the results obt,ained with glucagon (29, 30), parathyroid hormone by guest on March 24, 2020 http://www.jbc.org/ Downloaded from 2408 Porcine Thyrocalcitonin Vol. 245,No. 9 (al), and adrenocorticotropic hormone (21). The polypeptide fragments of the S-peptide of ribonuclease (31), and the tetradecapeptide from myoglobin (32) have also been shown to contain little helical structure in aqueous solution.
TC, however, does not appear to possess a rigid conformation and a coil ti helix equilibrium may exist in water with guanidine shifting the equilibrium in favor of the coil and 2-chloroethanol in favor of the helical form. Similarly, the structure of glucagon has been shown to vary depending on its environment, and is helical in the crystalline state (29,30,33). These results suggest that TC, as well as other polypeptide hormones, do not exist in rigid conformations in aqueous solutions, and that the polypeptides are capable of forming helical structures in nonpolar solvents. This type of transition may occur in lipid layers or cellular membranes at their receptor sites where the water concentration is significantly reduced.
Both human and salmon TC are significantly different in sequence and have a longer biological half-life when compared to the porcine hormone. It will be of interest to compare the physicochemical properties of TC from a variety of species in order to gain a greater appreciatioii of the relationships between protein structure and physiological function.