Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Prosaposin is a 517-amino-acid protein that, through proteolytic processing, gives rise to four sphingolipid activator proteins, saposin A, B, C, and D (O'Brien and Kishimoto, 1991). In addition, prosaposin is secreted in human milk, seminal fluid, cerebrospinal fluid, and serum (O'Brien and Kishimoto, 1991). Prosaposin is present in high concentrations in the rat (Kondoh et al., 1993), mouse (Kreda et al., 1994; Sun et al., 1994), and human (O'Brien et al., 1988) nervous systems, is concentrated within neuronal cell membranes (Fu et al., 1994), and is secreted after sciatic nerve injury (Hiraiwa et al., 1999). Prosaposin has been shown to have neurotrophic factor activity (O'Brien et al., 1994; Qi et al., 1996) and this activity was localized to a 14-amino-acid region of the saposin C domain. Subsequently, a series of Prosaptide (trademark of Myelos Corporation) peptides have been synthesized that retain this activity (O'Brien et al., 1995; Taylor et al., 2000). Prosaptide peptides have been shown to stimulate neurite outgrowth, choline acetyltransferase activity (O'Brien et al., 1995; Kotani et al., 1996a; Qi et al., 1996, 1999), and prevent apoptosis of cerebellar granule neurons (Tsuboi et al., 1998) and Schwann cells (Campana et al., 1999). In the peripheral nervous system, administration of Prosaptide peptides facilitated sciatic nerve regeneration (Kotani et al., 1996b), prevented paclitaxel-induced peripheral thermal hypoalgesia (Campana et al., 1998a), and improved diabetic neuropathy in rats (Calcutt et al., 1997, 1999). Prosaposin and Prosaptide peptides also prevented neuronal death and the associated learning disabilities caused by cerebral ischemia (Sano et al., 1994; Kotani et al. 1996a; Igase et al., 1999) or by a cortical stab wound (Hozumi et al., 1999).

In our effort to develop Prosaptide peptides for the systemic treatment of neurodegenerative diseases, we have investigated the blood-brain permeability of selected Prosaptide peptides and their stability in vivo. We previously reported that Prosaptide TX14(A) crossed the blood-brain barrier, however, it was rapidly degraded in both serum and brain (Taylor et al., 2000). A second peptide, Prosaptide TX15-2, crossed the blood-brain barrier and had increased stability in brain (Taylor et al., 2000).

Recent developments in peptidomimetics, based on peptide-bond reversal and inversion of chirality (Chorev and Goodman, 1993, 1995), have presented an increased possibility of designing superior peptide-based therapeutics. Here, we have demonstrated that retro-inverso Prosaptide peptidomimetics (peptides in which the primary sequence is reversed and D- rather than L-amino acids are used) retain neurotrophic factor activity. One of these Prosaptide peptidomimetics, Prosaptide D4, remained intact in both serum and brain and was transported across the blood-brain barrier. To our knowledge, this is the first report of a bioactive retro-inverso neurotrophic factor.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Animals. Male Sprague-Dawley rats (250-300 g) were provided by Harlan Industries (San Diego, CA). All animal experiments were conducted in accordance with the National Institutes of Health Guide for the Care and Use of Laboratory Animals.

Prosaptides. Anaspec (San Jose, CA) provided Prosaptide peptides at greater than 95% purity. The sequence of all peptides used in these studies is presented in Table 1. Prosaptide D4 was iodinated with iodobeads (Pierce, Rockford, IL) and Na¹²⁵I (NEN, Boston, MA) with 0.25 mCi/12.5 µg of peptide, using an 8-min labeling time. Iodinated Prosaptide D4 was purified from unincorporated iodine with a 5-ml Sephadex G-10 column equilibrated in PBS. The radiopurity of iodinated Prosaptide D4 was assessed by HPLC. The specific activity of iodinated peptides was 5.5 to 6.5 mCi/mg. Human serum albumin, ^{99 M}Tc were purchased from MediPhysics (San Diego, CA), and the albumin labeled according to the manufacturer's instructions.

Neurite Outgrowth. Cell culture reagents were purchased from Life Technologies (Grand Island, NY). The mouse neuroblastoma cell line NS20Y was a generous gift from Drs. T. Taketomi and K. Uemura (Shinshi University, Matsumoto, Japan). Cells were maintained as previously described (Taylor et al., 2000).

Stimulation of Guanosine-5'-O-(3-thio)triphosphate (GTPS) Binding. SHSY5Y cells were the generous gift of Dr. Stephen Fisher (University of Michigan, Ann Arbor). The assay was as described previously (Thomas et al., 1995; Hiraiwa et al., 1997). SHSY5Y cell membrane preparations (50-100 µg of protein) were incubated with 125 µCi of [³⁵S]GTPS (1250 Ci/nmol; NEN). GDP (3 µM) was added to amplify the difference between ligand-stimulated and background binding. Unlabeled GTPS (10 nM) was also added to define nonspecific binding and this value was subtracted from specific binding. All assays were performed in duplicate.

Blood-Brain Barrier Transport. Transport experiments were conducted according to the methods of Patlak et al. (1983) and Blasberg et al. (1983), as adapted by Banks et al. (1993). Rats were anesthetized using 65 mg/kg sodium pentobarbital (Abbott Laboratories, Santa Clara, CA) and the right jugular vein and left carotid artery exposed. Ten million cpm of ¹²⁵I-Prosaptide D4 and 2 × 10⁶ cpm ^{99 M}Tc-albumin (plasma marker) were injected together into the jugular vein in a volume of 100 µl of PBS. At various times, serum and brain were collected and radioactivity measured using a gamma-counter. The brain/blood ratio (ml/g) for ¹²⁵I and ^{99 M}Tc and exposure times were calculated as described by Patlak et al. (1983) using the equation brain/blood ratio = dpm in brain/g divided by dpm in serum/ml. The amount of ¹²⁵I present in brain was corrected for the amount of serum present in brain (¹²⁵I  ^{99 M}Tc) and the corrected ¹²⁵I brain/blood ratios were plotted against exposure time. The rate constant for unidirectional influx (K_i, in units of ml · g¹ · min¹) of Prosaptide D4 was taken from the slope of the line.

In Vivo Stability. To assess the in vivo stability of Prosaptide D4, ¹²⁵I-Prosaptide D4 was administered as described above. Serum and brain were collected at various times after injection. Prosaptide D4 was extracted from brain by homogenization in 5 ml of ice-cold extraction solution [56% acetonitrile in 0.1% trifluoracetic acid (TFA) containing 10 mM each of N-ethylmaleimide; 1,10-phenanthroline; EDTA; and D-thyroxine]. Homogenates were centrifuged and supernatants dried overnight. Dried supernatants were resuspended in 0.1% TFA, filtered, and applied to a C8 reversed phase HPLC column. Fractions were collected and radioactivity counted. Serum was simply dried down, resuspended in 0.1% TFA, and applied to the C8 column. Processing controls were used to measure the amount of degradation that occurred during processing and were prepared by adding ¹²⁵I-Prosaptide D4 to blood or brain in vitro and then processing them as described above. In vivo degradation results were then corrected by dividing them by the value for processing degradation.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

A schematic representation of prosaposin, the amino acid sequence of saposin C, and the sequence and location of the neurotrophic region of prosaposin are shown in Fig. 1. Table 1 shows the wild-type sequence of the neurotrophic region within saposin C, the sequences of TX14(A) and TX15-2, two L-amino acid Prosaptide peptides previously examined for stability and blood-brain barrier permeability (Taylor et al., 2000), and the sequences of Prosaptides D1-D5 that have been examined in this study. Prosaptide D1, containing all D-amino acids in a standard orientation, lacked biological activity in the neurite outgrowth assay (Table 1). In contrast, the retro-inverso Prosaptides D2-D5 were biologically active (Table 1). In a neurite outgrowth assay, each of these retro-inverso Prosaptides had an ED₅₀ between 0.2 and 0.8 nM. Prosaptide D5 had an ED₅₀ of 0.2 nM, which was 4 times more active than the most active L-amino acid Prosaptide TX14(A) (Table 1).

View larger version (10K): [in this window] [in a new window]   Fig. 1.   Schematic representation of prosaposin, the sequence of human saposin C, and the neurotrophic region it contains. The design of all Prosaptide peptides is based on this 14-amino-acid sequence.

All of the retro-inverso Prosaptides stimulated [³⁵S]GTPS binding to SHSY5Y cell membranes (Fig. 2) and stimulation was 45 to 60% of the control. Prosaptide 14 M1 (TKLIDNDKTEKEIL), an inactive L-amino acid Prosaptide peptide, inhibited the action of Prosaptide D5.

View larger version (49K): [in this window] [in a new window]   Fig. 2.   Activity of Prosaptides D1-D5 in [35S]GTPS binding assay. Cell membranes were prepared from SKNMC cells and incubated with 10 ng/ml Prosaptide peptides in the presence of [35S]GTP. The amount of binding on stimulation with peptides was measured. Data shown are the mean ± S.E. (n = 3).

The in vivo stability and blood-brain barrier permeability of Prosaptide D4 were examined. This retro-inverso peptide was chosen for further examination because it contains a tyrosine and was readily labeled with ¹²⁵I. Prosaptide D4 remained intact in serum or brain during the 60-min time period examined (Fig. 3) with 102 ± 1.1 and 99 ± 0.4% of the radioactivity obtained from serum and brain homogenates eluting in the position of intact ¹²⁵I-Prosaptide D4, respectively. Prosaptide D4 was rapidly cleared from serum with a t_1/2 of 3.3 min (Fig. 4). Labeled Prosaptide D4 was transported into brain and had a unidirectional influx rate constant (K_i) of 2.5 × 10⁴ ml · g¹ · min¹ (Fig. 5).

View larger version (13K): [in this window] [in a new window]   Fig. 3.   In vivo stability of Prosaptide D4 in serum () and brain (). Iodinated peptides were injected i.v. At various times after injection serum and brain were taken and processed for HPLC analysis. The percentage of intact was calculated from the elution profiles of radioactivity. Each point is data from one animal. Because there is no trend for degradation over time and the result is a flat line (serum: y = 0.037x + 101; r2 = 0.23; brain: y = 0.013x + 99; r2 = 0.19), it is legitimate to combine the data points over time and thus n = 4.

View larger version (11K): [in this window] [in a new window]   Fig. 4.   Clearance of Prosaptide D4 from serum. At various times after injection, radioactivity was measured in serum and its log value plotted against time. All data from three separate experiments is shown (n = 25). A line was fitted to the curve (y = 9.2x + 5.1, r2 = 0.94; n = 15) and t1/2 of disappearance from serum calculated from the inverse of the slope × 0.301.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Transfer of Prosaptide D4 across the blood-brain barrier. Iodinated peptides were injected i.v. At various times after injection, the radioactivity was measured in serum and brain. The brain/blood ratio of radioactivity was plotted against the exposure time and a line was fitted to the data (y = 2.5 × 104x + 1.4, r2 = 0.65; n = 15). The Ki of 2.5 × 104 ml · g1 · min1 was taken from the slope of the line. All data from three separate experiments are shown (n = 25).

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

In this article, we have described the design of retro-inverso Prosaptide peptidomimetics that retain bioactivity. In general, bioactive retro-inverso peptidomimetics have increased bioactivity compared with the native structure (for review, see Chorev and Goodman, 1993, 1995). This was true for Prosaptides D2-D5. The ED₅₀ for Prosaptide D5 was 0.2 nM, which is 7 times lower than that of the wild-type Prosaptide peptide. In some cases, retro-inverso peptidomimetics can mimic the antigenic and receptor binding properties of the parent peptide (Guichard et al., 1996). Like wild-type Prosaptide, Prosaptide TX14(A) and Prosaptide TX15-2, Prosaptides D2-D5 not only stimulated neurite outgrowth but also stimulated the binding of [³⁵S]GTPS to cell membranes. Thus, retro-inverso Prosaptide peptidomimetics appear to bind to the same receptor as native prosaposin and L-amino acid Prosaptide peptides (Campana et al., 1996, 1998b; Hiraiwa et al., 1997). The finding that Prosaptide 14 M1, an inactive mutant variant of Prosaptide TX14(A), attenuated the [³⁵S]GTPS binding stimulated by Prosaptide D5 supports the idea of specific receptor competition. These data also define Prosaptide14 M1 as an antagonist of the putative prosaposin receptor.

The folding of a retro-inverso peptide may be hindered when the peptide contains a helical segment (Guichard et al., 1996). Data indicate that the antigenic and receptor ligand mimicry between L-amino acid and retro-inverso peptides occurs only when the retro-inverso peptide is in random coil loop or cyclic conformations. The neurotrophic segments in prosaposin can be mimicked by a cyclic peptide, flanked by helical stretches (Liepinsh et al., 1997). The ends of the neurotrophic segment are near each other, with approximately 7 Å between C atoms of residues 17 and 30. A loop conformation of the retro-inverso Prosaptides may be the basis for the bioactivity of D2-D5 peptides. This is yet to be confirmed. Missing from this model is the role of the N-linked oligosaccharide chain at threonine 24 in saposin C (Ito et al., 1993). The retro-inverso Prosaptide D5 is very active and yet does not include an oligosaccharide chain. However, the natural glycopeptide ligand may differ in several aspects from the synthesized peptides such as binding to a lectin site or protection from proteases. Thus, it is fortuitous that retro-inverso Prosaptides are highly bioactive as neurotrophic factors. Further investigation of the structural elements that contribute to the bioactivity of Prosaptide D2-D5 is underway.

Retro-inverso peptides, in which all D-amino acids are used and the change in chirality is counteracted by reversing the primary sequence, contain inter-amino acid bonds that are the most closely related isosteric replacements for the original peptide bond. These modifications preserve the major structural characteristics of the peptide backbone while substantially changing the native structure. Consequently, retro-inverso peptides generally have increased stability as has been demonstrated for a number of peptides, including enkephalin, glutathione, Substance P, gastrin, and atrial natiuretic peptide (for review, see Chorev and Goodman, 1993, 1995). Prosaptide D4 was stable in vivo with no degradation observed over the 60 min examined. This is in contrast to both Prosaptide TX14(A) and Prosaptide TX15-2. We previously demonstrated that at 60 min after i.v. administration, 30% of radioactivity was present as intact Prosaptide TX14(A) in serum and 0% in brain. Prosaptide TX5-2 had increased brain stability and yet only 50% of radioactivity was detectable as intact Prosaptide TX15-2 at 60 min. There was no detectable Prosaptide TX15-2 in serum at that time. Thus, the use of retro-inversion greatly increased the stability of a Prosaptide peptide.

Prosaptide D4 crossed the blood-brain barrier. However, the rate of transfer (K_i) was approximately 10-fold lower than that for Prosaptide TX14(A) and Prosaptide TX15-2 (Taylor et al., 2000). If Prosaptide D4 is transported across the blood-brain barrier by simple diffusion, then, given its molecular weight and the fact that diffusion across the blood-brain barrier is proportional to the inverse of the square of the molecular weight, Prosaptide D4 would be expected to cross the blood-brain barrier with a rate of 10⁵ to 10⁴ ml · g¹ · min¹. The results obtained are well within this range and support the idea that Prosaptide D4 crosses the blood-brain barrier by simple diffusion (Banks et al., 1985). Other substances with similar transport rates that have central nervous system effects after systemic administration include leptin (Banks et al., 1996), morphine (Banks et al., 1994), insulin (Banks et al., 1997), and interleukin-1 (Banks et al., 1989).

In conclusion, we report the design of bioactive retro-inverso Prosaptides peptidomimetics. One of these peptidomimetics remains intact in vivo and crosses the blood-brain barrier. Thus, retro-inverso Prosaptide peptidomimetics may be useful in the therapy of central nervous system disorders.