Leukocyte antigen-related protein tyrosine phosphatase receptor: a small ectodomain isoform functions as a homophilic ligand and promotes neurite outgrowth

Abstract

The identities of ligands interacting with protein tyrosine phosphatase (PTP) receptors to regulate neurite outgrowth remain mainly unknown. Analysis of cDNA and genomic clones encoding the rat leukocyte common antigen-related (LAR) PTP receptor predicted a small, approximately 11 kDa ectodomain isoform, designated LARFN5C, containing a novel N terminal followed by a C-terminal segment of the LAR fifth fibronectin type III domain. RT-PCR and Northern blot analysis confirmed the presence of LARFN5C transcripts in brain. Transfection of COS cells with LARFN5C-Fc cDNA resulted in expression of the predicted protein, and Western blot analysis verified expression of approximately 11 kDa LARFN5C protein in vivo and its developmental regulation. Beads coated with rLARFN5C demonstrated aggregation consistent with homophilic binding, and pull-down and immunoprecipitation assays demonstrated that rLARFN5C associates with the LAR receptor. rLARFN5C binding to COS cells was dependent on LAR expression, and rLARFN5C binding to LAR +/+ hippocampal neurons was fivefold greater than that found by using LAR-deficient (-/-) neurons. Substratum-bound rLARFN5C had potent neurite-promoting effects on LAR +/+ neurons, with a fivefold loss in potency with the use of LAR -/- neurons. rLARFN5C in solution at low nanomolar concentrations inhibited neurite outgrowth induced by substratum-bound rLARFN5C, consistent with receptor-based function. These studies suggest that a small ectodomain isoform of a PTP receptor can function as a ligand for the same receptor to promote neurite outgrowth.

Fig. 1.

Genomic organization and alternative splicing predicting the LARFN5C isoform. A, Full-length LAR contains three IgG domains, eight FNIII domains, a transmembrane (TM) domain, and two PTP catalytic domains (D1, D2). LASE-c is a nine residue insert introduced into the FNIII-5 domain by alternative splicing. B, C, Genomic and cDNA sequence analysis demonstrated the LASE-e retained intron immediately upstream of the exon encoding LASE-c and the LASE-d retained intron immediately downstream of the exon encoding the C-terminal half of the FNIII-5 domain (FNIII-5C). The most downstream in-frame ATG start site in LASE-e and the upstream most in-frame TGA stop site in LASE-d are shown and predict an ∼11 kDa isoform consisting of a 49 residue N terminal followed by the LASE-c and FNIII-5C domains. Constitutive LAR transcript splicing is indicated by the bottom thicker splice indicator line, and alternative splicing is shown by the top thinner lines.D, Genomic sequence demonstrates the splice donor (gt underlined) and splice acceptor (ag underlined) sites of the ∼2.7 kb LASE-e retained intron and the ∼1.3 kb intron separating the LASE-c and FNIII-5C exons. Exonic sequence is shown in uppercase letters and intronic in lower case letters; LARFN5C novel exonic sequence is shown in lower case letters with amino acid translation; stop codons are boxed. Antibody LARFN5C 29–48 was raised against a synthetic peptide corresponding to N-terminal residues PGPLQAKPFTILSPFLSSRC.

Fig. 2.

RT-PCR, Northern blot, and Western blot analysis of LARFN5C expression. A, RT-PCR analysis was performed by using an upstream primer corresponding to a site within LASE-e and a downstream primer corresponding to a site within LASE-d generating a predicted product of 534 bp. Poly(A) mRNA (0.05 μg/reaction) from rat whole brain and rat E16 hippocampus was assessed. H₂O served as a negative control. B, Northern blot analysis of P21 hippocampal poly(A) RNA was conducted by using a riboprobe corresponding to the 150 nucleotides encoding the novel N terminal of LARFN5C (P1) or corresponding to the 147 nucleotides encoding the LASE-c and FNIII-5 regions of LARFN5C (P2). P1 detected an ∼1.6 kb transcript, and P2 detected an ∼1.6 kb transcript of identical size as well as the expected LAR ∼8 kb transcript. C, Northern blot autoradiograms derived by using the P2 probe and RNA isolated from P21 and adult (Ad) hippocampus were assessed by scanning densitometry; the ratio of LARFN5C (∼1.6 kb) over the LAR (∼8 kb) signal was calculated (mean ± SE; n = 6 blots). A small but significant decrease in this ratio was detected in the adult when compared with the P21 samples (p < 0.05, Mann–Whitney).D, Protein extract (40 μg) from E16 mouse hippocampal tissue was applied to each lane. Blots were incubated with the indicated antibodies. Negative controls included omission of primary antibody, use of preimmune antiserum, and preincubation of primary antibody with LARFN5C 29–48 peptide. LARFN5C 29–48 and LASE-c antibodies detected an ∼11 kDa protein (bottom arrow) consistent with the predicted size of LARFN5C. Both antibodies also detected an ∼44 kDa signal (top arrow). E, E16 hippocampal tissue extracts were immunoprecipitated (IP) with preimmune, anti-LARFN5C 29–48, or LASE-c antibodies. Western blot analysis of IPs with anti-LARFN5C 29–48 or LASE-c antibodies detected ∼11 and ∼44 kDa proteins. F, Protein extract (40 μg) was applied to each lane at the indicated developmental time points from E16 hippocampal tissue and E18 cultured hippocampal neurons derived from LAR +/+ and −/− mice and hippocampal tissues. Blots were incubated with LARFN5C 29–48 antibody followed by reprobing with actin antibody (bottom panel). Similar levels of LARFN5C were present in tissue and cells derived from LAR +/+ and −/− mice. G, Western blot autoradiograms derived by using LARFN5C 29–48 antibody were assessed by scanning densitometry. The ratio of LARFN5C signal over the actin signal demonstrated that LARFN5C protein levels in the hippocampus decreased significantly during development by 94% (mean ratios ± SE; p < 0.05; n= 4; ANOVA).

Fig. 3.

Purification of rLARFN5C and homophilic binding of rLARFN5C. A, Left, His-tagged recombinant LARFN5C (rLARFN5C) protein was eluted from metal affinity resin and applied to a 4–12% gradient polyacrylamide gel system. Staining with GelCode Blue reagent detected an ∼16 kDa protein along with a faint signal at ∼64 kDa. A, Right, Analysis that used the more sensitive LASE-c antibody Western blot detected signal at ∼16, ∼32, and ∼64 kDa. B, Left, Incubation of His-tagged rLARFN5C with enterokinase generated the predicted ∼11 kDa product as detected by GelCode Blue staining. B, Right, Analysis that used the more sensitive LARFN5C 29–48 antibody Western blots detected signal at ∼11, ∼22, and ∼44 kDa. C, Fluorescent microspheres were coated with either BSA (top two panels) or rLARFN5C (remainder of panels), sonicated for 30 sec, incubated for 1 hr without or with the indicated antibodies, and then examined under fluorescent microscopy. BSA-coated microspheres remained dissociated, whereas rLARFN5C-coated microspheres in the absence of antibody or in the presence of preimmune antibodies underwent aggregation. LARFN5C 29–48 or LASE-c antibodies blocked aggregation.

Fig. 4.

LAR Western blot, rLARFN5C pull-down, LARFN5C immunoprecipitation, and rLARFN5C-Fc expression assays.A, Western blot analysis that used the LAR N-terminal antibody assessed LAR protein expression in extracts prepared from E18 hippocampal cultures derived from LAR +/+ and −/− mice. Blots were reprobed with β-actin antibody to control for differences in loading (bottom panel). In LAR −/− extracts only trace levels of the LAR ∼150 kDa (top arrow) and ∼110 kDa (middle arrow) isoforms were present, whereas similar levels of the ∼80 kDa (bottom arrow) isoform were present in LAR −/− and +/+ samples. B, For pull-down assays E18 hippocampal extracts were incubated with metal affinity resin coated with His-tagged rLARFN5C or His-tagged Positope control protein. After incubation the resin was washed, and bound proteins were eluted and assessed via Western blots with LAR N-terminal antibody. Incubation of extracts with His-tagged Positope failed to pull down protein detected by the LAR antibody (lanes 1, 2). Incubation of LAR +/+ extracts with His-tagged rLARFN5C resulted in the capture of ∼150 kDa (top arrow), ∼110 kDa (middle arrow), and ∼80 kDa (bottom arrow) proteins detected by LAR antibody (lane 3). Incubation with LAR −/− extract failed to detect the ∼150 and ∼110 kDa isoforms (lane 4), consistent with their relative absence in LAR −/− tissue. C, E16 hippocampus lysates were immunoprecipitated with preimmune or LARFN5C 29–48 antibody, and immune complexes were analyzed by Western blotting with the use of LAR N terminus monoclonal antibody. LAR ∼150 kDa (top arrow) and ∼110 kDa (bottom arrow) isoforms were found to coimmunoprecipitate with LARFN5C, but not with preimmune antibody. The bottom band at ∼55 kDa is consistent with nonspecific IgG binding. D, COS cells were transfected with pcDNA3.1 vector in the null form or containing the LARFN5C-Fc insert. Western blot analysis of cell pellet extract and culture medium collected at 48 hr, which used Fc fragment-specific antibody, detected the expected ∼42 kDa LARFN5C-Fc fusion protein.

Fig. 5.

LAR is required for rLARFN5C binding to COS cells.A, COS 7 cells were transfected with LAR (top row) or with null vector (bottom row). After transfection the cells were incubated with rLARFN5C recombinant protein (625 nm) followed by washes, immunostaining for LAR and LARFN5C, and confocal imaging. Column one shows COS cell morphology under differential interference contrast (DIC), columns two and three demonstrate imaging for LAR by using Cy3 (red) and for LARFN5C by using Alexa 488 (green), and column four shows image overlay. LARFN5C binding is entirely dependent on LAR expression; LARFN5C signal is colocalized mainly with LAR (yellow), whereas a portion of LAR signal (red) remains independent of LARFN5C. Scale bar, 10 μm. B, LARFN5C and LAR signal were measured for individual cells, and ratios were calculated over the indicated concentrations of rLARFN5C (mean ± SE;n = 75).

Fig. 6.

Exogenous rLARFN5C binding to high and low LAR-expressing neurons. A, E18 mouse hippocampal neurons derived from LAR +/+ and −/− embryos were combined and cocultured for 24 hr and coimmunostained with monoclonal LAR N-terminal antibody and LARFN5C 29–48 antibody. As expected in cocultures of LAR +/+ and −/− cells, populations of high LAR-expressing and low LAR-expressing cells were evident (top vs bottom panel). LAR and endogenous LARFN5C signal is present in cell bodies, along neurites, and in growth cones (arrowheads). Similar to COS cells, LARFN5C signal is colocalized mainly with LAR (yellow), whereas a portion of LAR signal (red) remains independent of LARFN5C. Similar levels of endogenous LARFN5C staining were observed in both high and low LAR-expressing cells. Signal overlay demonstrated minimal LAR–LARFN5C signal (yellow) in low LAR-expressing cells. Scale bar, 10 μm. B, Quantitation of LAR signal demonstrated distinct populations of cells: high LAR-expressing and low LAR-expressing. C, The addition of exogenous rLARFN5C to culture media resulted in increased LARFN5C signal associated with high LAR-expressing neurons as compared with low LAR-expressing neurons. Scale bar, 5 μm. D, Quantification of LAR and LARFN5C signal demonstrated a fivefold decrease in LAR signal in low LAR-expressing neurons and similar levels of endogenous LARFN5C signal in high and low LAR-expressing neurons (mean ± SE;n = 30; ***p < 0.001, Mann–Whitney test). E, In cultures containing added rLARFN5C at 625 nm, the quantification of LAR and LARFN5C signal demonstrated fivefold less LARFN5C signal associated with low LAR-expressing neurons as compared with that associated with high LAR-expressing neurons (mean ± SE; n = 30; ***p < 0.001, Mann–Whitney test).

Fig. 7.

Characterization of rLARFN5C neurite-promoting activity. E18 hippocampal neurons were cultured for 24 hr on nitrocellulose coated with BSA (A), laminin (B), rLARFN5C (C), or rLARFN5C (D) subjected to His tag immunoprecipitation. After fixation the neurons were immunostained with GAP-43 antibody, followed by Alexa 488 green-conjugated goat anti-rabbit IgG. E, E18 hippocampal neurons were plated on nitrocellulose coated with the indicated proteins (LN, laminin; FN, fibronectin). rLARFN5C was added to substrate either directly (rLARFN5C) or after pretreatment via immunoprecipitation (IP) with His tag or LASE-c antibody. BSA was included as an additional negative control. After 2 hr of incubation the cells were washed and fixed. The mean number of cells (±SE) present per field is shown. For laminin, fibronectin, and rLARFN5C ∼200–300 fields over a series of six separate studies were assessed. For each of the three negative control conditions ∼50 fields were assessed in total. F, E18 hippocampal neurons were plated on nitrocellulose coated with the indicated proteins. rLARFN5C was added to substrate either directly (rLARFN5C) or after pretreatment via IP with His tag or LASE-c antibody. After 24 hr the neurite lengths were measured (mean ± SE). For laminin, fibronectin, and rLARFN5C ∼400–900 neurites over a series of eight studies were measured. For each of the three negative control conditions 10–60 neurites were measured. G, Cumulative distribution of neurite length measurements (BSA, open circles; His tag IP, open boxes; LASE-c IP, open triangles; FN, filled triangles; rLARFN5C, filled circles; LN, filled boxes).H, E18 hippocampal neurons were plated on nitrocellulose substrate coated with the indicated concentrations of rLARFN5C. For each dose that was tested 80–250 neurites were measured. Mean neurite length ± SE is shown. Filled circles indicate results with ∼16 kDa uncleaved rLARFN5C, and open circles indicate results with enterokinase-cleaved ∼11 kDa rLARFN5C.I, E18 hippocampal neurons were plated on nitrocellulose substrate coated with 3.125 pmol of rLARFN5C, and soluble rLARFN5C was added to culture medium at the indicated concentrations. For each dose that was tested 70–200 neurites were measured. Mean neurite length ± SE is shown.

Fig. 8.

rLARFN5C-induced neurite outgrowth is partially dependent on the presence of neuronal LAR. A, E18 hippocampal neurons derived from LAR +/+ (filled triangles) and LAR −/− (open triangles) mice were cultured in wells coated with laminin applied at the indicated doses. After 24 hr the neurite lengths were determined (mean ± SE; 80–230 neurites were measured for each genotype and each concentration). The potency of laminin in inducing neurite outgrowth from LAR −/− neurons was unchanged.B, E18 hippocampal neurons derived from LAR +/+ and LAR −/− mice were cultured in wells coated with the rLARFN5C protein applied at the indicated doses. After 24 hr the neurite lengths were determined (mean ± SE; 70–220 neurites were measured for each genotype and each concentration). rLARFN5C had a fivefold loss of potency in LAR −/− cultures.