Function of three cytochromes in photosynthesis of whole cells of Rhodospirillum rubrum as studied by flash spectroscopy. Evidence for two types of reaction center.

1. Changes in the absorption spectrum induced by 10-mus flashes and continuous light of various intensities were studied in whole cells of Rhodospirillum rubrum in the presence and absence of 2-n-heptyl-4-hydroxyquinoline-N-oxide(HOQNO) and antimycin A. 2. Three cytochromes, c-420 (cytochrome c2), c-560 (cytochrome b) and c-428 were photoactive and gamma and alpha peaks at 420 and 550, 428 and 560, and 428 and 551 nm, respectively; they were photooxidized following the flash with half times of 0.3, 0.6 and 7 ms in the approximate ratios of 1/100, 1/300 and 1/1000 (cytochrome oxidized/antenna chlorophyll) and became reduced with half times of 12 ms, 60 ms and 0.7 s, respectively. c-428 and c-560 have not been distinguished before. 3. From a detailed analysis of the kinetics of P+ (oxidized reaction center chlorophyll) and the cytochromes, we conclude that 5% of the P+ (P2+) oxidizes c-428, whereas the remaining 95% of P+ (P1+) oxidizes c-420. At actinic light intensities low enough to keep c-420 fully reduced, approx. 4-5% of P becomes oxidized, accompanied by all c-428. The P2+ -P2 difference spectrum induced by this weak light is, when corrected for a shift to longer wavelengths of the bacteriochlorophyll absorption band at 878 nm, identical to the difference spectrum caused by the photooxidation of the remaining P1. At low flash intensity, c-428 becomes preferentially photooxidized, which suggests that the reaction centers where c-428 functions as a secondary donor contain much more antenna pigments compared to the centers where c-420 serves this purpose. 4. c+-420 is reduced in a competitive way by reduced c-560 (t 1/2=7 ms), and by an electron donor pool, (t 1/2=15 ms). HOQNO inhibits both pathways; antimycin A only the first. In the presence of HOQNO, c-560 is in the oxidized state in the dark, and is reduced in a light flash (t 1/2=100 ms), indicating that c-560 acts in a cyclic electron transport chain connected to P1. 5. The ratio of numbers of molecules P1 and antenna bacteriochlorophyll, transferring excitation energy to P1, is P1/bacteriochlorophyll1=1/30 P2: bacteriochlorophyll2=1/300; c-420/P1=1:2; c-560/P1=1/6; C-428/P2=1/1; bacteriochlorophyll2=7:3. If P2 is oxidized, excitation energy is transferred from bacteriochlorophyll2 to bacteriochlorophyll1.     

Function and properties of a soluble c-type cytochrome c-551 in secondary photosynthetic electron transport in whole cells of Chromatium vinosum as studied with flash spectroscopy

Changes in the absorption spectrum induced by 10-μs flashes and continuous light of various intensities were studied in whole cells of Chromatium vinosum.This paper describes the role and function of a soluble c-type cytochrome, c-551, which was surprisingly found to act in many ways similar to the cytochrome c-420 in Rhodospirillum rubrum, described in a previous paper [1].After the photooxidation of the membrane bound high potential cytochrome c-555 by a 10-μs flash, (the low potential cytochrome c-552 was kept permanently in the oxidized state) the oxidation of c-551 is observed ($\text{t}{}_{\mathrm{<mtext>1</mtext><mtext>2</mtext>}}\text{= 0.3}\text{ms}$). From a careful analysis of the absorbance difference spectrum and the kinetics it is concluded that there is approximately 0.6–0.7 c-551 per reaction center and that essentially all the c⁺-555 is reduced via the cytochrome c-551. The oxidized-reduced difference spectrum of c-551 shows peaks at 551 and 421.5 nm. The reduction of c⁺-551 following the flash-induced oxidation is strongly inhibited by HOQNO, but only slightly by antimycin A.Cytochrome c-551 reduces only the oxidized high potential cytochrome c-555, which is probably located on the outside of the membrane, on the opposite side of the primary acceptor. The low potential cytochrome c-552 does not show any detectable interaction with cytochrome c-551. After the cells have been sonicated, no c-551 is photooxidized and at least part of the cytochrome occurs in the solution.Analysis of the reduction kinetics of c⁺-551 in the absence and presence of external donors suggests that c⁺-551 is partly reduced via a cyclic pathway, which is blocked by addition of o-phenanthroline, and partly via a non-cyclic pathway. The non-cyclic reduction rate of c⁺-551 (k = 6 s^?1) is increased approximately 5–10 times upon thiosulphate addition, suggesting a role for c-551 between the final donor pool and the oxidized membrane bound c-type cytochromes.     

Rotational mobility of the photoreaction center in chromatophore membranes of Rhodospirillum rubrum

We investigated the rotational mobility of the photoreaction center in chromatophores of Rhodospirillum rubrum by studying the photoinduced linear dichroism of absorption changes at 865 nm. The study was carried out in suspensions of chromatophores treated with ferricyanide in order to bleach their antenna bacteriochlorophyll and thus minimize depolarization by energy transfer. Very little depolarization of the photoinduced absorbance change at 865 nm was observed at room temperature for chromatophores immersed in a highly viscous medium over the time range 0–10 ms following an exciting light flash. In the light of independent evidence for transmembrane arrangement of the photoreaction center, we conclude that the photoreaction center protein is immobilized in the chromatophore membrane for at least 10 ms.     

Energy and electron transfer in the photosynthetic reaction center complex of Acidiphilium rubrum containing Zn-bacteriochlorophyll a studied by femtosecond up-conversion spectroscopy

A photosynthetic reaction center (RC) complex was isolated from a purple bacterium, Acidiphilium rubrum. The RC contains bacteriochlorophyll a containing Zn as a central metal (Zn-BChl a) and bacteriopheophytin a (BPhe a) but no Mg-BChl a. The absorption peaks of the Zn-BChl a dimer (P_Zn), the accessory Zn-BChl a (B_Zn), and BPhe a (H) at 4 K in the RC showed peaks at 875, 792, and 753 nm, respectively. These peaks were shorter than the corresponding peaks in Rhodobacter sphaeroides RC that has Mg-BChl a. The kinetics of fluorescence from P_Zn^*, measured by fluorescence up-conversion, showed the rise and the major decay with time constants of 0.16 and 3.3 ps, respectively. The former represents the energy transfer from B_Zn^* to P_Zn, and the latter, the electron transfer from P_Zn to H. The angle between the transition dipoles of B_Zn and P_Zn was estimated to be 36° based on the fluorescence anisotropy. The time constants and the angle are almost equal to those in the Rb. sphaeroides RC. The high efficiency of A. rubrum RC seems to be enabled by the chemical property of Zn-BChl a and by the L168HE modification of the RC protein that modifies P_Zn.     

Necessity of a membrane component for nitrogenase activity in Rhodospirillum rubrum

Acetylene reduction catalyzed by nitrogenase from Rhodospirillum rubrum has low activity and exhibits a lag phase. The activity can be increased by the addition of a chromatophore membrane component and the lag eliminated by preincubation with this component, which can be solubilized from chromatophores by treatment with NaCl. It is both trypsin- and oxygen-sensitive. Titration of the membrane component with nitrogenase and vice versa shows a saturation point. The membrane component interacts specifically with the Fe protein of nitrogenase, the interaction being ATP- and Mg²⁺-dependent.     

Photooxidase activity of Rhodospirillum rubrum chromatophores and reaction center complexes. The role of non-cyclic electron transfer in generation of the membrane potential

The mechanism of light-induced O₂ uptake by chromatophores and isolated P-870 reaction center complexes from Rhodospirillum rubrum has been investigated.The process is inhibited by o-phenanthroline and also by an extraction of loosely bound quinones from chromatophores. Vitamin K-3 restored the o-phenanthroline-sensitive light-induced O₂ uptake by the extracted chromatophores and stimulated the O₂ uptake by the reaction center complexes. It is believed that photooxidase activity of native chromatophores is due to an interaction of loosely bound photoreduced ubiquinone with O₂. Another component distinguishable from the loosely bound ubiquinone is also oxidized by O₂ upon the addition of detergents (lauryldimethylamine oxide or Triton X-100) to the illuminated reaction center complexes and to the extracted or native chromatophores treated by o-phenanthroline. Two types of photooxidase activity are distinguished by their dependence on pH.The oxidation of chromatophore redox chain components due to photooxidase activity as well as the over-reduction of these components in chromatophores, incubated with 2,3,5,6-tetramethyl-p-phenylenediamine (Me₄Ph(NH₂)₂) or N,N,N′,N′-tetramethyl-p-phenylenediamine (TMPD) (plus ascorbate) in the absence of exogenous electron acceptors, leads to an inhibition of the membrane potential generation, as measured by the light-induced uptake of penetrating phenyldicarbaundecaborane anions (PCB^?) and tetraphenylborate anions. The inhibition of the penetrating anion responses observed under reducing conditions is removed by oxygen, 1,4-naphthoquinone, fumarate, vitamin K-3 and methylviologen, but not by NAD⁺ or benzylviologen. Since methylviologen does not act as an electron acceptor with the extracted chromatophores, it is believed that this compound, together with fumarate and O₂, gains electrons at the level of the loosely bound ubiquinone. Data on the relationship between photooxidase activity and membrane potential generation by the chromatophores show that non-cyclic electron transfer from reduced Me₄Ph(NH₂)₂ to the exogenous acceptors is an electrogenic process, whereas non-cyclic electron transfer from reduced TMPD is non-electrogenic.Being oxidized, Me₄Ph(NH₂)₂ and TMPD are capable of the shunting of the cyclic redox chain of the chromatophores. Experiments with extracted chromatophores show that the mechanisms of the shunting by Me₄Ph(NH₂)₂ and TMPD are different.     

Regulation of cyclic photophosphorylation in Rhodospirillum rubrum by the redox state of nicotinamide-adenine dinucleotide

We have investigated the effect of the redox state of added NAD on the rates of anaerobic cyclic photophosphorylation which are supported by membrane vesicles isolated from Rhodospirillum rubrum. As the redox potential of NAD was lowered, the activity decreased according to a typical potentiometric titration. The Nernst plot showed an apparent midpoint potential (E′_o) of ?350 mV and had a slope which corresponded to a two-electron transition. Besides, an almost identical potentiometric relationship was found to exist between the extent of light-elicited ATP formation in anaerobic suspensions of intact R. rubrum cells and the redox potential of intracellular NAD. These results suggest that physiological photophosphorylation in R. rubrum requires the oxidized form of a membrane-bound constituent (E′_o = ?350 mV) whose redox state is controlled by the redox state of cytoplasmic NAD.     

Photooxidase system of Rhodospirillum rubrum. I. Photooxidations catalyzed by chromatophores isolated from a mutant deficient in photooxidase activity

The aerobic photooxidations of reduced 2,6-dichlorophenolindophenol and of reaction-center bacteriochlorophyll (P-870) have been investigated in membrane vesicles (chromatophores) isolated from a non-phototrophic Rhodospirillum rubrum strain. In aerobic suspensions of wild-type chromatophores, continuous light elicits an increase of the levels of 2,6-dichlorophenolindophenol and of oxidized P-870, which reach steady-state values shortly after the onset of illumination. In contrast, light induces in mutant suspensions a transient increase of the levels of 2,6-dichlorophenolindophenol and of oxidized P-870, which fall to low steady-state values within a few seconds. These observations suggest that the mutation has altered a redox constituent located on the low-potential side of the photochemical reaction center, between a pool of acceptors and oxygen.Since endogenous cyclic photophosphorylation is catalyzed by mutant chromatophores at normal rates, it appears that the constituent altered by the mutation does not belong to the cyclic electron-transfer chain responsible for photophosphorylation. However, the system which mediates the aerobic photooxidations and the cyclic system are not completely independent: endogenous photophosphorylation is inhibited by oxygen in wild-type chromatophores but not in mutant chromatophores; in addition, the inhibitor of cyclic electron flow, 2-heptyl-4-hydroxyquinoline-N-oxide, enhances the aerobic photooxidation of reduced 2,6-dichlorophenolindophenol by chromatophores from both strains.These results support a tentative branched model for light-driven electron transfer. In that model, the constituent altered in the mutant strain is located in a side electron-transfer chain which connects the cyclic acceptors to oxygen.     

Inactivation of Rhodospirillum rubrum coupling factor by 7-chloro-4-nitrobenzo-2-oxa-1,3-diazole. Modification of a tyrosine protected by phosphate

Chemical modification of Rhodospirillum rubrum chromatophores by 7-chloro-4-nitrobenzo-2-oxa-1,3-diazole (NBD-Cl) results in inactivation of photophosphorylation, Mg²⁺-ATPase, oxidative phosphorylation and ATP-driven transhydrogenase, with apparent first-order kinetics. Other energy-linked reactions such as light-driven transhydrogenase and light-dependent proton uptake were insensitive to NBD-Cl. The Ca²⁺-ATPase activity of the soluble coupling factor from chromatophores (R. rubrum F₁) was inactivated by NBD-Cl with kinetics resembling those described for Mg²⁺-ATPase and photophosphorylation activities of chromatophores. Both NBD-chromatophores and NBD-R. rubrum F₁ fully recovered their activities when subjected to thiolysis by dithioerythritol. Phosphoryl transfer reactions of chromatophores and Ca²⁺-ATPase activity of R. rubrum F₁ were fully protected by 5 mM P_i against modification by NBD-Cl. ADP or ATP afforded partial protection. Analysis of the protection of Ca²⁺-ATPase activity by P_i indicated that NBD-Cl and P_i are mutually exclusive ligands. Spectroscopic studies revealed that tyrosine and sulfhydryl residues in R. rubrum F₁ underwent modification by NBD-Cl. However, the inactivation was only related to the modification of tyrosine groups.     

Triplet formation on a monomeric chlorophyll in the photosystem II reaction center as studied by time-resolved infrared spectroscopy

The process of formation of the triplet state of chlorophyll in the photosystem II (PS II) reaction center complex was studied by means of time-resolved infrared (IR) spectroscopy. Using a dispersive-type IR spectrometer with a time resolution of approximately 55 ns, transient spectra in the C=O stretching region (1760--1600 cm(-1)) were measured at 77 K. The data were analyzed by singular-value decomposition and subsequent least-squares fitting. Two distinct spectral components having different kinetic behaviors were resolved. One had spectral features characterized by negative peaks at 1740 and 1680 cm(-1) and an overall positive background and was assigned to the P(680)(+)Phe(-)/P(680)Phe radical pair by static FTIR measurements of the P(680)(+)/P(680) and Phe(-)/Phe differences. The other had prominent negative and positive peaks at 1668 and 1628 cm(-1), respectively, which were previously assigned to the keto C==O change upon triplet formation of the monomeric chlorophyll denoted as Chl(T) [Noguchi, T., Tomo, T., and Inoue, Y. (1998) Biochemistry 37, 13614-13625]. The former component of P(680)(+)Phe(-)/P(680)Phe exhibited a multiphasic decay with time constants of 77 ns (75%), 640 ns (18%), 8.3 micros (4%), and 0.3 ms (3%), while the latter component of (3)Chl(T)/Chl(T) was formed with a single-exponential rise with a time constant of 57 ns and had a lifetime of 1.5 ms. From the observations that only the two spectral components were resolved without any other triplet intermediates and the time constant of (3)Chl(T) formation roughly agreed with or seemed even faster than that of the major phase of the P(680)(+)Phe(-) decay, two alternative mechanisms of triplet formation are proposed. (i) (3)Chl(T) is directly formed from P(680)(+)Phe(-) by charge recombination at Chl(T), and (ii) (3)P(680) is formed, and then the triplet is transferred to Chl(T) with a time constant of much less than 50 ns. The location of Chl(T) in the D1 subunit as the monomer chlorophyll corresponding to the accessory bacteriochlorophyll in the L subunit of purple bacteria is favored to explain the former mechanism as well as the triplet properties reported in the literature. The physiological role of the triplet formation on Chl(T) is also discussed.     

Affinity chromatography of H+-translocating adenosine triphosphatase isolated by chloroform extraction of Rhodospirillum rubrum chromatophores. Modification of binding affinity by divalent cations and activating anions

1. ATPase isolated from Rhodospirillum rubrum by chloroform extraction and purified by gel filtration or affinity chromatography shows three bands (α, β and γ) upon electrophoresis in sodium dodecyl sulphate.2. Ca²⁺-ATPase activity of the preparation is inhibited by aurovertin and efrapeptin but not by oligomycin. Activity may be inhibited by treatment with 4-chloro-7-nitrobenzofurazan and subsequently restored by dithiothreitol.3. The enzyme fails to reconstitute photophosphorylation in chromatophores depleted of ATPase by sonic irradiation.4. Most of the active protein from the crude chloroform extract binds to an affinity chromatography column bearing an immobilised ADP analogue but not to a column bearing immobilised pyrophosphate.5. In the absence of divalent cations, a component with a very high specific activity for Ca²⁺-ATPase is eluted from the column by 1.6 mM ATP. This protein migrates as a single band on 5% polyacrylamide gel electrophoresis and only possesses three subunits. At 12 mM ATP an inactive protein is eluted which does not run on acid or alkali polyacrylamide gels and shows a complex subunit structure.6. ATPase preparations prepared by acetone extraction or by sonic irradiation of chromatophores may also be purified 10-fold by affinity chromatography.7. The inclusion of 5 mM MgCl₂ or CaCl₂ during affinity chromatography of chloroform ATPase increases the capacity of the column for the enzyme and demands a higher eluting concentration of ATP.8. When the enzyme is more than 90% inhibited by efrapeptin or 4-chloro-7-nitrobenzofurazan, the binding characteristics of the enzyme are not affected.9. 10 mM Na₂SO₃, which greatly stimulates the Ca²⁺- and Mg²⁺-dependent ATPase activity of the enzyme and increases K_i (ADP) for Ca²⁺-ATPase from 50 to 850 μM, prevents binding to the affinity column. Binding may be restored by the addition of divalent cations.10. Na₂SO₃ increases the rate of ATP hydrolysis, ATP-driven H⁺ translocation and ATP-driven transhydrogenase in chromatophores.11. It is proposed that anions such as sulphite convert the chromatophore ATPase into a form which is a more efficient energy transducer.     

Magnetic field-stimulated luminescence and a matrix model for energy transfer. A new method for determining the redox state of the first quinone acceptor in the reaction center of whole cells of Rhodospirillum rubrum

In whole cells of Rhodospirillum rubrum the light-induced absorbance difference spectrum of the reduction of the first quinone electron acceptor Q₁ was determined in order to relate the emission yield ф and the magnetic field-induced emission increase Δф to the redox state of Q₁. It was found that Δф/ф² is a linear function of the number of reaction centers, in which Q₁ is reduced, independent of the fraction of reaction centers in the oxidized state. The emission yield is a hyperbolic function of the fraction of reaction centers closed, either by reduction of the acceptor Q₁ or by oxidation of the primary electron donor P. Apparently, in whole cells of R. rubrum a matrix model for energy transfer between various photosynthetic units can be applied. A model is presented, which is a generalization of theoretical considerations reported before (Duysens, L.N.M. (1978) in Chlorophyll Organization and Energy Transfer in Photosynthesis, Ciba Found. Symp. 61 (New Series), pp. 323–340, Elsevier/North-Holland, Amsterdam) and which is in excellent agreement with the experiments. From simultaneous measurements of Δф and ф the redox state of the reaction center can relatively easily be determined. So far, this is the only method for simultaneously measuring the fractions P⁺ and Q⁻₁ in intact cells under steady-state conditions.     

Solvent damping of internal processes in myoglobin studied by specific heat spectroscopy and flash photolysis.

We address the question of dynamic coupling between protein and solvent by comparing the enthalpy relaxation of the solvent (75% v/v glycerol-water) to internal ligand binding in myoglobin. When the solvent relaxation is slow compared to intramolecular events we observe decoupling of protein motions from the solvent. In the opposite limit there is a significant contribution of the solvent to internal friction. The solvent enhances the apparent activation energy of transitions in myoglobin. This result is discussed in terms of a generalized Kramer's law involving a dynamic friction coefficient.     

The kinetics and thermodynamics of the reaction of solid-state fully reduced membrane-bound cytochrome oxidase with carbon monoxide as studied by dual-wavelength multichannel spectroscopy and flash photolysis.

1. The results of non-linear optimization studies on the mechanism of reaction of solid-state fully reduced membrane-bound cytochrome oxidase with CO over the 178--203 K range are presented. The analysis is carried out on data obtained by dual-wavelength multichannel spectroscopy at three wavelength pairs (444--463 nm, 590--630 nm and 608--630 nm), which yield three distinct progress curves. The only model that satisfies the triple requirement of a standard deviation within the standard error of the data, a random distribution of residuals and good determination of the optimized parameters is a two-species sequential mechanism: flash photolysis yields unliganded cytochrome oxidase and free CO, which then recombine to form species Ic; Ic is then converted into species IIc, which is identical with the cytochrome oxidase-CO complex existing before flash photolysis. All the thermodynamic parameters describing this model are calculated. 2. On the basis of the data obtained from this paper, together with data from potentiometric studies, magnetic susceptibility measurements and i.r. spectroscopy, the chemical identity of the species is suggested.     

Regulation of melanin synthesis in mammalian cells, as studied by somatic hybridization. I. Evidence for negative control

Somatic hybrids between pigmented Syrian hamster cells and unpigmented mouse cells were isolated and propagated in vitro. These hybrids are unpigmented and lack dopa oxidase (and tyrosinase) activity, which is correlated with the pigmentation of the Syrian hamster cells. In contrast, the presence of three other enzymes (LDH, MDH, and thymidine kinase) also specified by the hamster genome but unrelated to pigment synthesis is observed in the hybrid cells. This suggests that the repression of dopa oxidase in these cells is a specific effect on the enzyme associated with the differentiated state of pigment cells. It is concluded that the genetic control of differentiation in this case involves a diffusible regulator substance which functions negatively.     

Energy and electron transfer in the photosynthetic reaction center complex of Acidiphilium rubrum containing Zn-bacteriochlorophyll a studied by femtosecond up-conversion spectroscopy

A photosynthetic reaction center (RC) complex was isolated from a purple bacterium, Acidiphilium rubrum. The RC contains bacteriochlorophyll a containing Zn as a central metal (Zn-BChl a) and bacteriopheophytin a (BPhe a) but no Mg-BChl a. The absorption peaks of the Zn-BChl a dimer (P(Zn)), the accessory Zn-BChl a (B(Zn)), and BPhe a (H) at 4 K in the RC showed peaks at 875, 792, and 753 nm, respectively. These peaks were shorter than the corresponding peaks in Rhodobacter sphaeroides RC that has Mg-BChl a. The kinetics of fluorescence from P(Zn)(*), measured by fluorescence up-conversion, showed the rise and the major decay with time constants of 0.16 and 3.3 ps, respectively. The former represents the energy transfer from B(Zn)(*) to P(Zn), and the latter, the electron transfer from P(Zn) to H. The angle between the transition dipoles of B(Zn) and P(Zn) was estimated to be 36 degrees based on the fluorescence anisotropy. The time constants and the angle are almost equal to those in the Rb. sphaeroides RC. The high efficiency of A. rubrum RC seems to be enabled by the chemical property of Zn-BChl a and by the L168HE modification of the RC protein that modifies P(Zn).     

Evidence for the existence of two types of cAMP binding sites in aggregating cells of Dictyostelium discoideum.

Inhibition by dithiothreitol of the cell-bound phosphodiesterase has allowed the measurement of cAMP binding to aggregation-competent Dictyostelium discoidium amebae. Two classes of binding sites were demonstrated: Type 1, of low affinity (Kd approximately 160 nM) and high capacity (Ro approximately 1 X 10(5) sites per cell); and Type 2, of high affinity (Kd approximately 9 nM) but low capacity (Ro approximately 1.5 X 10(4) sites per cell). Both sites are developmentally regulated and are expressed during aggregation. The specificities of both sites are consistent with the specificity observed in vivo for the chemotactic response.     

Parathyroid hormone-related protein. Evidence for secretion of a novel mid-region fragment by three different cell types.

The cDNA-predicted amino acid sequence of parathyroid hormone-related protein (PTHrP) contains multiple basic amino acid motifs, suggesting that PTHrP undergoes extensive post-translational processing prior to secretion. The secretory forms of the peptide are currently unknown. To identify these secretory forms, medium was harvested from three cell types: human renal carcinoma (SKRC-1) cells, human keratinocytes, and rat insulinoma cells stably transfected with the cDNA for PTHrP(1-141) (RIN-141 cells). Amino-terminal species were immunopurified using an anti-PTHrP(1-36) column, and mid-region species using an anti-PTHrP(37-74) column. PTHrP peptides in medium and in cell extracts were further resolved by reverse phase high performance liquid chromatography (RP-HPLC) and identified using region-specific immunoassays. SKRC-1 and RIN-141 cells secreted three distinct amino-terminal species and a novel, non-amino-terminal, mid-region fragment. Sequence and sodium dodecyl sulfate-polyacrylamide gel electrophoresis analysis indicated that the RIN-141 cell mid-region fragment begins at amino acid 38 of the cDNA-predicted sequence and is approximately 70 amino acids in length. Comparison of RP-HPLC elution patterns suggests that SKRC-1 cells and keratinocytes secrete a similar or identical mid-region fragment. Immunofluorescence studies revealed a Golgi pattern for the amino-terminal species and a secretory granule pattern for the mid-region fragment. These studies indicate that 1) multiple PTHrP species are secreted, including a novel mid-region fragment; 2) Arg37 serves as a cleavage site in at least three cell types; 3) PTHrP(1-36) is likely to be an authentic secretory form of PTHrP; and 4) the mid-region fragment appears to be packaged into secretory granules. The marked interspecies conservation of this mid-region PTHrP suggests that it will have important biological functions.