Isolation and characterisation of arthropod gap junctions

Abstract

Gap junctions have been isolated from the hepatopancreas of the crustacean arthropod, Nephrops norvegicus (Norway lobster). SDS‐PAGE of these preparations shows two major protein bands, mol. wt. 18 000 (18 K) and mol. wt. 28 000 (28 K). The 18‐K and 28‐K proteins are interconvertible, cannot be distinguished by two dimensional tryptic and chymotryptic peptide mapping, and therefore appear to be different (most likely monomeric and dimeric) forms of the same protein. The protein can also aggregate to higher multimeric forms mol. wt. 38 000 (presumed trimer), and mol. wt. 52 000 (presumed tetramer). The buoyant density of the isolated gap junctions in continuous potassium iodide gradients is 1.260 g/cm3. The junctions are progressively solubilized in increasing SDS concentrations, mostly between 0.1% and 0.2% SDS, and this is accompanied by the release of the 18‐K and 28‐K forms of the junctional protein. The Nephrops hepatopancreas 18‐K junctional protein has antigenic determinants in common with the vertebrate 16‐K junctional protein as shown by cross‐reactivity with two different affinity purified antibody preparations. However, no detectable similarity can be seen between the major 125I‐labelled tryptic and chymotrytpic peptides of the Nephrops hepatopancreas 18‐K protein and the mouse liver 16‐K protein.

A 16 kDa protein co‐isolating with gap junctions from brain tissue belonging to the class of proteolipids of the vacuolar H+‐ATPases

“…First, from SDS-PAGE analysis it follows that in both procedures a 16 kDa polypeptide co-isolates with the plasma membrane fraction consisting of gap junction-like plaques under EM. Secondly, a comparison of the amino acid sequence with an unpublished sequence of a 16 kDa protein from liver isolated according to the Finbow technique [6] (Hertzberg, personal communication) revealed a high degree of homology differing only in positions 5 and 15 to our sequence.…”

“…gap junct., 16 kDa polypeptide from bovine brain gap junction preparations; chrom.gr., 16 kDa proteolipid from bovine vacuolar H+-ATPase from chromaffin granules [8]; yeast vac., homologous sequence deduced from a reading frame from Saccharomyces cerevisiae [15]; boy. Fo9, 8 kDa proteolipid from bovine mitochondria ATPase [16]; yeast F09, 8 kDa proteolipid from yeast mitochondria ATPase [16]; CFoIII, 8 kDa proteolipid from spinach chloroplast.ATP synthase [16] By the detergent treatment with 1~70 Triton X-100 on complete tissue homogenates instead of on isolated plasma membrane fractions, a 16 kDa protein was consistently obtained [6]. This bore no relationship to the connexin family from biochemical and immunochemical results.…”

“…In addition, a 16 kDa protein isolated from different species and phyla has frequently been propounded as the major structural component of gap junctions [6,7]. gap junct., 16 kDa polypeptide from bovine brain gap junction preparations; chrom.gr., 16 kDa proteolipid from bovine vacuolar H+-ATPase from chromaffin granules [8]; yeast vac., homologous sequence deduced from a reading frame from Saccharomyces cerevisiae [15]; boy.…”

“…This bore no relationship to the connexin family from biochemical and immunochemical results. Although we used a different approach in isolating the 16 kDa protein from brain, two main reasons point to the identical nature of this protein to that described earlier [6]. First, from SDS-PAGE analysis it follows that in both procedures a 16 kDa polypeptide co-isolates with the plasma membrane fraction consisting of gap junction-like plaques under EM.…”

“…These proteins, here referred to as connexins (cx) first suggested by Goodenough [3] vary in molecular masses between 26 kDa (cx26) [5], 32 kDa (cx32) [1,2] and 43 kDa (cx43) [3] based on their predicted amino acid sequence from their encoding cDNAs. A 16 kDa protein frequently reported in diverse tissues [6] is thought to represent a gap junction constituent. However, the nature of this protein remains obscure since homology based on immunochemical criteria and two-dimensional pep-tide mapping [7] on connexins have not been evidenced.…”