The alga Volvox carteri represents one of the simplest multicellular organisms. Its extracellular matrix (ECM) is modified under developmental control, e.g. under the influence of the sex‐inducing pheromone that triggers development of males and females at a concentration below 10−16 M. A novel ECM glycoprotein (pherophorin‐S) synthesized in response to this pheromone was identified and characterized. Although being a typical member of the pherophorins, which are identified by a C‐terminal domain with sequence homology to the sex‐inducing pheromone, pherophorin‐S exhibits a completely novel set of properties. In contrast to the other members of the family, which are found as part of the insoluble ECM structures of the cellular zone, pherophorin‐S is targeted to the cell‐free interior of the spherical organism and remains in a soluble state. A main structural difference is the presence of a polyhydroxyproline spacer in pherophorin‐S that is linked to a saccharide containing a phosphodiester bridge between two arabinose residues. Sequence comparisons indicate that the self‐assembling proteins that create the main parts of the complex Volvox ECM have evolved from a common ancestral gene.
Some of the simplest multicellular organisms are found among the green algae of the genus Volvox. Volvox carteri is composed of only two cell types: 2000–4000 biflagellate Chlamydomonas‐like somatic cells are arranged in a monolayer at the surface of a hollow sphere and 16 much larger reproductive cells (‘gonidia’) lie just below the somatic cell sheet (Starr, 1969). Due to its simplicity, Volvox is an ideal model for a biochemical analysis of developmental processes.
Many developmental responses of cells are mediated by the extracellular matrix (ECM) with which those cells are in contact. In Volvox, this is very clearly demonstrated in the fascinating process of sexual differentiation triggered by the sex‐inducing pheromone. Volvox cells are surrounded and held together by a glycoprotein‐rich ECM (for a review see Kirk et al., 1986). The Volvox ECM shows a distinct structural architecture. The outermost area, called the boundary zone (BZ; for nomenclature see Figure 1A) (Kirk et al., 1986), contains those components of the ECM that appear to be continuous over the surface of the organism. The area lying internal to the boundary zone, called the cellular zone (CZ), exhibits specializations around individual cells. The deep zone (DZ) of the ECM consists of components that fill the cell‐free interior of the spherical organism. This zone contains highly viscous polysaccharide‐rich amorphous components (Kirk et al., 1986).
The chemical composition of the Volvox ECM is strongly modified under the influence of the sex‐inducing pheromone. The pheromone, a glycoprotein (Starr and Jaenicke, 1974; Tschochner et al., 1987; Mages et al., 1988), converts asexually growing males and females to the sexual pathway. This pheromone is among the most potent biological effector molecules known: it exhibits full effectiveness at 6×10−17 M (Starr, 1970; Gilles et al., 1984). Many lines of evidence indicate that the ECM plays a key role in this sexual induction process. The earliest biochemical responses to the pheromone detected so far are structural modifications within the ECM (Wenzl and Sumper, 1982, 1986b; Gilles et al., 1983). The CZ of the Volvox ECM contains members of the newly described pherophorin family. Pherophorins are glycoproteins that contain a C‐terminal domain with homology to the sexual pheromone. Pherophorin I is constitutively expressed and represents a main component of the cellular zone of the ECM. Under the influence of the pheromone, synthesis of pherophorin II is initiated and its C‐terminal domain becomes proteolytically liberated from the parent glycoprotein (Sumper et al., 1993). It has been proposed that this modification of the ECM is part of a signal amplification process required to obtain the exquisite sensitivity of this sexual induction system. So far, the DZ has not been investigated in detail for any biochemical changes in response to the sexual pheromone, although some early observations indicate a modification of this ECM region (Gilles et al., 1983).
In this paper, we describe a novel pherophorin that is synthesized in response to the sex‐inducing pheromone. Although highly homologous to other members of the family that are located exclusively within the somatic cell sheets, this novel pherophorin (pherophorin‐S) is specifically targeted to the DZ of the ECM. Pherophorin–S exhibits a unique glycosylation pattern among the pherophorins: it contains a phosphodiester bridge between two arabinose residues.
Identification of pherophorin‐S
Pheromone‐induced changes in the composition of the ECM have been characterized in detail only within the CZ (Figure 1A) of the ECM (Wenzl and Sumper, 1982, 1986b; Ertl et al., 1989; Sumper et al., 1993; Godl et al., 1995). In order to extend this analysis to the DZ (Figure 1A) of the ECM, which may constitute >90% of the total volume of the organism, the composition of this ECM compartment from asexual and sexually induced organisms was compared after pulse labelling with radioactive sulfate. Mild mechanical stress as may be exerted by forcing Volvox spheroids through a hypodermic needle fragments the spheroids, producing hemispheres or smaller fragments of cellular sheets. The material of the DZ is thereby selectively released. This mild fragmentation of the organism does not affect viability of the cells. After low speed centrifugation, the cell‐free supernatant, containing the material from the DZ was subjected to SDS–PAGE. About 30 min after application of the sexual pheromone, synthesis of a previously unobserved 35S‐labelled component, with an apparent molecular mass of 110 kDa, becomes detectable in the fluorogram of the SDS–polyacrylamide gel (Figure 1B, DZ, ind). This component is synthesized only transiently. Maximum expression is found 120 min after application of the pheromone. For reasons explained below, this component of the DZ was named pherophorin‐S. Pherophorin‐S is only detectable in the DZ of the ECM and is not detectable in asexually growing organisms. As pherophorin‐S is quantitively extracted without any additives (detergents, salt or EDTA), it is a soluble component of the ECM.
In contrast to other members of the pherophorin family which are proteolytically processed (Sumper et al., 1993; Godl et al., 1995), pherophorin‐S is a stable protein, as demonstrated by a pulse–chase labelling experiment (Figure 1B).
Pulse labelling experiments using radioactive phosphate revealed that, in contrast to all other known members of the pherophorin family that have been studied previously, pherophorin‐S also incorporates phosphate. The chemical nature of the incorporated phosphate will be described below.
Purification of pherophorin‐S
Mild mechanical disruption of Volvox spheroids liberates the material of the DZ, including pherophorin‐S. This property was used to purify pherophorin‐S; the corresponding extract will be denoted in the following as ‘deep zone extract’. After filtration and centrifugation, the colourless extract containing pherophorin‐S was fractionated by anion exchange chromatography (Q‐Sepharose followed by Mono Q). Final purification was achieved by preparative SDS–PAGE.
Purified pherophorin‐S exhibits different apparent molecular masses on SDS–PAGE, depending on the percentage of the acrylamide used. The observed values range from 90 to 130 kDa (∼100 kDa on an 8% gel). This is a property of some glycosylated proteins. Treatment of pherophorin‐S with anhydrous hydrogen fluoride at 0°C, a procedure that selectively deglycosylates glycoproteins but does not cleave polypeptides (Mort and Lamport, 1977), reduces the apparent molecular mass by ∼10 kDa (8% SDS–PAGE). Therefore, pherophorin‐S is a glycoprotein.
To obtain amino acid sequence data, purified pherophorin‐S glycoprotein was digested with trypsin and the resulting peptide mixture was separated by reversed phase C2/C18 HPLC. The material of well‐separated peaks was directly subjected to amino acid sequence analysis on an automated gas phase sequencer. The amino acid sequence data obtained are underlined in Figure 2B.
Cloning of the pherophorin‐S gene
The amino acid sequence of the tryptic peptide IYPSVGSSSIVTPSWTAIGG was used to synthesize an antisense oligonucleotide primer to reverse transcribe mRNA isolated from sexually induced Volvox algae. A sense primer derived from the same peptide allowed amplification by PCR of a cDNA probe of 56 bp in length (Figure 2A, probe), which was cloned into the SmaI site of pUC18 by blunt ligation. Sequencing of this insert revealed a nucleotide sequence coding for the amino acid sequence of the peptide mentioned above. The RACE‐PCR technique (Frohman et al., 1988) was used to obtain additional sequence information. A 3′‐RACE‐PCR produced the 3′‐end of the mRNA, whereas successive 5′‐RACE‐PCRs did not arrive at the 5′‐end of the mRNA (Figure 2A). Similar problems observed with hydroxyproline‐rich glycoproteins from Volvox (Ertl et al., 1989, 1992) suggested that the pherophorin‐S mRNA might also contain a C‐rich stretch that causes premature termination of reverse transcription. Thus a genomic library of V.carteri constructed in the replacement vector λEMBL3 was screened to obtain the missing sequence data from a genomic clone. Nine positive clones were identified out of 60 000 phages screened. The ∼17 kb insert of one of these clones was subcloned and sequenced. A GC‐rich section coding for a proline‐rich domain was indeed identified. The sequence upstream of this GC‐rich section was then established by two additional 5′‐RACE‐PCRs (Figure 2A), each resulting in the same 5′‐end. Comparison of genomic and cDNA sequences revealed the presence of two introns. The strategy applied to collect the complete cDNA sequence is schematically summarized in Figure 2A. The deduced amino acid sequence for pherophorin‐S is given in Figure 2B. A molecular mass of 63.4 kDa was calculated for the polypeptide chain of pherophorin‐S. This is much less than the apparent molecular mass seen on SDS–polyacrylamide gels (Figure 1B, DZ, ind). The amino acid sequence of pherophorin‐S exhibits five N‐glycosylation sites, but glycosylation only accounts for an increase in molecular mass of ∼10 kDa (8% SDS–PAGE). Most likely, the presence of a domain with a very high proline/hydroxyproline content explains the difference between the observed and calculated molecular masses of deglycosylated pherophorin‐S, because such a domain has a reduced ability to bind SDS (Andres et al., 1993).
Pherophorin‐S is a member of the pherophorin family
A striking feature of the deduced amino acid sequence is a central domain, 88 amino acid residues in length, that is composed almost exclusively (89%) of proline residues. Most probably, the secondary structure of this domain is a polyproline II helix that separates the N‐ and C‐terminal domains. A BLASTP search (Altschul et al., 1990) of the SwissProt Protein Sequence Database revealed significant identities of the deduced amino acid sequence to the pherophorin family from Volvox (Figure 3). The region of identity covers nearly the total length of the polypeptide chain. For instance, the N‐terminal part of the polypeptide (amino acids 42–207) shows 55.3% identity to the N‐terminal part of pherophorin II (amino acids 1–168) (Sumper et al., 1993), the C‐terminal half (amino acids 308–596) exhibits 40.4% identity to the C‐terminal half of pherophorin II (amino acids 186–484). The name pherophorin‐S was chosen to indicate the fact that this polypeptide is the first member of the pherophorin family remaining highly soluble within the ECM. Pherophorin‐S contains 29 cysteines, most of them conserved among the different pherophorins, indicating conservation of the three‐dimensional structure.
Although pherophorin‐S is without any doubt a member of the pherophorin family, its chemical properties are strikingly different from all the other members known so far. First, pherophorin‐S is targeted to the DZ compartment of the ECM, whereas pherophorins I–III are located within the CZ of the ECM. Second, all other pherophorins are highly insoluble components of the ECM and, third, only pherophorin‐S incorporates phosphate.
It is the proline‐rich domain that is completely different in pherophorin‐S: this domain is ∼88 amino acids in length, whereas corresponding elements in pherophorins I–III are only ∼10 amino acids in length (Figure 3). As demonstrated below, it is this polyproline‐domain of pherophorin‐S that incorporates phosphate.
The promoter of the pherophorin‐S gene mediates pheromone‐dependent transcription
To examine regulation of the pherophorin‐S gene, the pherophorin‐S 5′‐untranslated region (∼1 kb) was placed in front of a reporter gene, the arylsulfatase gene from Volvox (Hallmann and Sumper, 1994b) (Figure 4A). In wild‐type Volvox, arylsulfatase is only expressed under sulfur starvation; no activity is detectable in organisms grown in sulfate‐containing medium (Hallmann and Sumper, 1994a).
After transformation of Volvox with the chimeric pherophorin‐S–arylsulfatase gene, the reverse transcription‐PCR technique was used to verify the existence of hybrid mRNA in transformants (see Materials and methods). Volvox transformants containing the arylsulphatase gene under the control of the pherophorin‐S promoter were incubated with or without the pheromone in the presence of the chromogenic enzyme substrate 4‐nitrocatechol sulfate. Arylsulfatase activity was determined photometrically by measuring the absorbance of the liberated 4‐nitrocatechol. Only transformants treated with the sex‐inducing pheromone exhibited enzyme activity (Figure 4B). Since arylsulfatase is extremly stable, activity could even be assayed in SDS–polyacrylamide gels using the chromogenic substrate 5‐bromo‐4‐chloro‐3‐indolyl sulfate (Figure 4C).
Thus, the ∼1 kb DNA fragment isolated from the upstream region of the phorophorin‐S gene mediates transcription of the arylsulphatase reporter gene in response to the sex‐inducing pheromone.
Transgenic Volvox expressing pherophorin‐S
To obtain sufficient amounts of pherophorin‐S for structural studies, transgenic Volvox were generated that express the pherophorin‐S gene under the control of the strong Volvox β‐tubulin promoter (Figure 5A). Stable transformants were produced as previously described (Hallmann and Sumper, 1994b, 1996; Schiedlmeier et al., 1994). The expression rate was 20‐ to 30‐fold higher in vegetatively grown Volvox transformants than in sexually induced wild‐type algae (Figure 5B). The transgenic Volvox strain did not have any visible phenotype.
The overexpression of pherophorin‐S in clone PheroS‐T1 allowed a much simpler purification protocol. Asexually grown PheroS‐T1 algae were disrupted. After centrifugation the supernatant was brought to 10% acetonitrile and passed over a C18 cartridge. The flow‐through was concentrated and applied to preparative SDS–PAGE. Pure pherophorin‐S could be eluted from the gel. Recombinant pherophorin‐S was used for all further chemical characterizations.
The characteristics of pherophorin‐S from asexually growing transformant PheroS‐T1 were compared with that obtained from sexually induced wild‐type algae. Neither targeting into the DZ of the ECM nor post‐translational modification causing incorporation of phosphate is affected in the transformant, indicating that these properties are not under the control of the sex‐inducing pheromone. The apparent molecular masses of pherophorin‐S from wild‐type algae and from transformant PheroS‐T1 are identical (Figure 5C).
Identification of a phosphodiester between arabinose residues in pherophorin‐S
The carbohydrate composition of pherophorin‐S was determinded by radio gas chromatography. Pherophorin–S purified from the transgenic Volvox strain grown in the presence of [14C]bicarbonate was hydrolysed and the resulting monosaccharides were converted to the corresponding alditol acetates. Pherophorin‐S contains the neutral sugars arabinose and galactose in a 1:1 ratio (Figure 6A).
Pherophorin‐S incorporates [33P]phosphate in pulse‐labelling experiments. Incorporated radioactivity is quantitatively removed from the polypeptide chain upon treatment with anhydrous HF, indicating that phosphate is not linked to a hydroxyamino acid. In addition, hydrolysis of pherophorin‐S in 0.5 M trifluoroacetic acid at 100°C for 2 h quantitatively liberates bound phosphate as a low molecular mass derivative. Analysis of this hydrolysate on polyethyleneimine thin‐layer plates resulted in the detection of two radioactive spots. One radioactive product migrated like a phosphomonoester, the main degradation product migrated like a phosphodiester (Figure 7A). The latter substance stained with orcinol reagent and was completely hydrolysed after 2 h in 6 M HCl at 100°C. Sugar analysis by gas chromatography identified arabinose as the only sugar present in this derivative (Figure 6C). In earlier studies, the phosphodiester arabinose‐5‐phospho‐5′‐arabinose was identified as a structural element in the ECM glycoprotein SSG 185 from Volvox (Holst et al., 1989). Proof for the existence of the same structural component in pherophorin‐S was obtained by mass spectrometry as follows. Pherophorin‐S (supplemented with trace amounts of labelled material) was hydrolysed and reduced. The resulting products were separated by high performance anion exchange chromatography (HPAEC) (Figure 7B). Radioactive fractions were subjected to thin‐layer chromatography (data not shown) and to mass spectrometry. After reduction of the anomeric carbon atoms with NaBH4, the phosphodiester produced a mass signal at 365.0 (Figure 7C). This exactly corresponds to the calculated mass for the reduced phosphodiester.
The polyproline domain carries the phosphodiester
Pherophorin‐S is not completely digested if treated with proteases like pronase, proteinase K or subtilisin. Rather, a resistant core with an apparent molecular mass of ∼50 kDa (8% SDS–PAGE) remains (Figure 8A). Proteolytic degradation of 33P‐labelled pherophorin‐S results in a core material that still contains all of the originally incorporated radioactivity. Consequently, the phosphodiester is located within this protease‐resistant core material. In order to define this core material, purified pherophorin‐S was digested with subtilisin. The 50 kDa core material was eluted from a preparative SDS–PAGE, deglycosylated with anhydrous HF (because the glycosylated material could not be analysed by Edman degradation) and purified by reversed phase HPLC (C2/C18). A single peptide eluted at 20% acetonitrile. Edman degradation of this material resulted in the sequence shown in Figure 8B. As expected, the resistant core material represents the proline‐rich domain of pherophorin‐S. The amino acid sequence analysis also confirms that the prolines at the very beginning of this domain (residues 211 and 213) remain unmodified, whereas prolines 215, 218, 219, 221 and 223 (and probably all the following) become hydroxylated and can serve as saccharide attachment sites.
The carbohydrate composition of the protease‐resistant 50 kDa core material was determined by radio gas chromatography. The core material contains the neutral sugars arabinose and galactose in a 1:1 ratio (Figure 6B), exactly as found for intact pherophorin‐S.
Sequence homology proves that pherophorin‐S is a member of the pherophorin family of Volvox ECM proteins. Like pherophorin II, it is synthesized in response to the sex‐inducing pheromone. However, pherophorin‐S exhibits unique properties: it is accumulated in the DZ of the Volvox ECM in a completely soluble state, in contrast to all the other members known so far, which are insoluble and restricted to the CZ. Biogenesis of ECMs occurs by self‐assembly, which means that each component contains within its structure the information necessary for this fascinating process. In pherophorins I–III the N‐ and C‐terminal domains are separated by a short polyhydroxyproline spacer. It is this spacer element that is strikingly different in pherophorin‐S: a stretch of ∼90 amino acid residues that are almost exclusively hydroxyproline residues separates the terminal domains. This spacer is glycosylated and, of particular interest, contains a phosphodiester bridge between two arabinose residues. This type of modification was originally discovered in another ECM glycoprotein, namely SSG 185 from Volvox (Ertl et al., 1989). SSG 185 is the monomeric precursor of a polymeric substructure within the CZ of the ECM that surrounds individual cells, creating honeycomb‐like chambers. Remarkably, SSG 185 contains exactly the same type of polyhydroxyproline spacer. Since SSG 185 and pherophorin‐S are found in completely different regions of the ECM, it is unlikely that this particular spacer provides the signal for targeting pherophorin‐S to the DZ. Using the newly established system of Volvox transformation (Schiedlmeier et al., 1994), chimeras of pherophorin domains should allow identification of structural elements that are responsible for specific targeting within the ECM.
SSG 185 and pherophorin I represent the main components of the cellular zone of the ECM in asexually growing Volvox (Godl et al., 1995). Under the influence of the sex‐inducing pheromone, pherophorin II is deposited within the CZ of the ECM and newly synthesized pherophorin‐S modifies the composition of the DZ. As the primary structures of all these ECM glycoproteins are known, it is possible to compare their modular composition and to search for structural homologies among the modules. Figure 9A presents the domain structure of these glycoproteins in diagrammatic fashion. The C‐terminal domains of all pherophorins (B‐type domain) are characterized by sequence homology with the sex‐inducing pheromone. The N‐terminal domains of all pherophorins (A–type domain) are related to both the N‐ and C‐terminal domains of SSG 185 (Figure 9B) (Godl et al., 1995). Moreover, a sequence comparison of the A‐ and B‐type domains of pherophorin II reveals that even these two regions exhibit 24% sequence identity over a stretch of 179 amino acid residues. Thus, the main parts of the complex ECM of Volvox, and even the species‐specific signalling molecule (sex‐inducing pheromone), appear to have been derived from the same ancestral gene (Figure 9B). Differently modulated domains have been linked together via spacer elements that are composed of polyhydroxyproline sequences and it is only the introduction of these spacers that qualifies these Volvox ECM proteins for membership in the class of hydroxyproline‐rich glycoproteins (HRPGs) typical of plant cell walls. These molecular data offer strong support for the idea of a gene superfamily of hydroxyproline‐rich glycoproteins (for reviews see Kieliszewski and Lamport, 1994; Woessner and Goodenough, 1994) from which new ECM proteins could constantly evolve. Volvocine radiation, i.e. the transition from unicellular Chlamydomonas to multicellular Volvox, is a recent event, probably having occurred within the past 50–75 million years (Rausch et al., 1989). This recent transition to multicellularity converted a simple cell wall into a complex ECM. Thus, this independent development of an ECM should offer an attractive model for studying the mechanisms operating during this process. The development of the Volvox ECM also appears to provide another example of the combinatorial advantage of shuffling modules, as is so evident in the evolution of the metazoan ECM (for a review see Doolittle, 1995).
Materials and methods
Wild‐type V.carteri strain HK10 (female) was obtained from the Culture Collection of Algae at the University of Texas (R.C.Starr). Mutant strain 153‐48 of V.carteri(Adams et al., 1990), obtained from D.L.Kirk (Washington University, St Louis, MO) was used as the DNA recipient in transformation experiments. This strain with wild‐type morphology carries a stable loss‐of‐function mutation in nitA, the structural gene encoding nitrate reductase (Gruber et al., 1996).
Synchronous cultures were grown in Volvox medium (Provasoli and Pintner, 1959) at 28°C in a 8 h dark/16 h light (10 000 lux) cycle (Starr and Jaenicke, 1974). The non‐selective medium used in transformation experiments was Volvox medium supplemented with 1 mM NH4Cl; selective medium was Volvox medium lacking NH4Cl and containing only nitrate as a nitrogen source.
Radioactive labelling of pherophorin‐S with [35S]sulfate or [33P]phosphate
Pulse labelling with [35S]sulfate was performed as described by Wenzl and Sumper (1981). In pulse and pulse‐chase experiments with [33P]phosphate, Volvox spheroids were washed thoroughly with and then suspended in 1 ml glycerophosphate‐free Volvox medium. After the addition of 50 μCi [33P]phosphate, incubation under standard conditions was continued for 0.5 or 1 h.
Purification of pherophorin‐S
Sexually induced wild‐type Volvox spheroids from six 20 l cultures were harvested by filtration on a 100 μm mesh nylon screen. The spheroids were broken up by forcing them through a 0.5 mm hypodermic needle. The disrupted spheroids were centrifuged at 20 000 g for 30 min. In order to remove any remaining insoluble components, the supernatant was brought to 25 mM Tris–HCl, pH 7.5, 0.9 M NaCl, 1 mM EDTA, 0.5 mM phenylmethylsulfonyl fluoride (PMSF) and applied to a QAE‐Sephadex A‐25 column (Pharmacia) equilibrated with the same buffer. Pherophorin‐S does not bind and was therefore detected in the flow‐through. After this filtration step, the material containing pherophorin‐S was diluted to 0.3 M NaCl and applied to a Q‐Sepharose FF anion exchange column (Pharmacia). Elution was performed with a linear gradient of 0.3–0.9 M NaCl in 25 mM Tris–HCl, pH 7.5, 1 mM EDTA, 0.5 mM PMSF. Pherophorin‐S elutes at ∼0.7 M NaCl. The fraction containing pherophorin‐S was diluted to 0.35 M NaCl and applied to a MonoQ HR 5/5 FPLC anion exchange column (Pharmacia). Again, elution was performed with a linear gradient of 0.35–1.5 M NaCl in 25 mM Tris–HCl, pH 7.5, 1 mM EDTA, 0.5 mM PMSF. Pherophorin‐S was recovered at ∼0.80–0.85 M NaCl. Fractions containing pherophorin‐S were concentrated by precipitation with deoxycholate and trichloroacetic acid (Mahuran et al., 1983). Final purification was achieved by preparative SDS–PAGE (7%).
Proteolytic digestion and separation of peptides
Aliquots of 20–30 μg pherophorin‐S were applied to a 7% SDS–PAGE gel and stained with Coomassie brilliant blue. The gel slice containing pherophorin‐S was cut into small pieces. Further treatment and digestion with trypsin was performed as decribed by Selmer et al. (1996). The resulting peptides were eluted from the gel by diffusion in 0.2 M (NH4)HCO3/50% acetonitrile. The eluate was passed through a 0.22 μm filter (Millipore), brought to 0.1% trifluoroacetic acid and dried by lyophilization. The peptides were dissolved in 6 M guanidine–HCl/ 0.1% trifluoroacetic acid and fractionated by reversed phase HPLC (SMART system; Pharmacia) on a 3 μm μRPC C2/C18 column (Pharmacia). Peptides were eluted with a 30 min linear gradient of 5–40% acetonitrile in 0.1% trifluoroacetic acid with a flow rate of 200 μl/min. Peptides were sequenced by Edman degradation using an automated gas phase peptide sequencer (Applied Biosystems, Foster City, CA).
Generation of a cDNA probe by PCR
Generation of a cDNA probe for pherophorin‐S was performed using the degenerate antisense oligonucleotide primer CCKATNGCNGTCCA and the degenerate sense oligonucleotide primer ATHTAYCCNAGYGT. The resulting 56 bp cDNA fragment was ligated into the SmaI site of vector pUC18 by blunt ligation and sequenced.
Cloning of the pherophorin‐S gene
The RACE‐PCRs were performed as described by Frohman et al. (1988). The V.carteri genomic library in λEMBL3 (Frischauf et al., 1983) described by Ertl et al. (1989) was used to clone the pherophorin‐S gene. The screening procedure followed standard techniques (Sambrook et al., 1989). DNA sequencing was performed by the chain termination method (Sanger et al., 1977) using T7 DNA polymerase (Pharmacia).
Construction of the chimeric pherophorin‐S–arylsulfatase gene
The Volvox arylsulfatase reporter gene (Hallmann and Sumper, 1994b) was placed under the control of the pherophorin‐S gene 5′‐region. Additional restriction sites were introduced by PCR to facilitate ligation of the parent DNAs. An EcoRV site was generated directly in front of the start codons of both the pherophorin‐S gene and the Volvox arylsulfatase gene. A KpnI site was introduced into the pherophorin‐S 5′‐region ∼1 kb upstream of the start codon. Then, a KpnI–EcoRV fragment covering ∼1 kb upstream sequence of the pherophorin‐S gene was ligated to an ∼10 kb EcoRV–SalI fragment containing the Volvox arylsulfatase gene (Hallmann and Sumper, 1994a) with its 15 introns. The complete construct was confirmed by sequencing.
Construction of the chimeric β‐tubulin–pherophorin‐S gene
To achieve high pherophorin‐S production, the pherophorin‐S gene was placed under the control of the Volvox β‐tubulin promoter. For construction of the chimeric gene, genomic clones of Volvox β‐tubulin (Harper and Mages, 1988) and pherophorin‐S were used. An additional EcoRV site was generated by PCR directly in front of the start codons of both the pherophorin‐S gene and the Volvox β‐tubulin gene to facilitate ligation of the parental DNAs. A KpnI site was introduced into the β‐tubulin promoter region ∼0.5 kb upstream of the start codon. Then, a KpnI–EcoRV fragment bearing the ∼0.5 kb Volvox β‐tubulin promoter region was ligated to an ∼9 kb EcoRV–SmaI fragment containing the pherophorin‐S gene. During cloning a T nucleotide from the EcoRV site in front of the start ATG was deleted for unknown reasons, destroying the EcoRV site. The complete construct was confirmed by sequencing.
Stable nuclear transformation of Volvox carteri
Volvox carteri strain 153‐48 was transformed by using a particle gun to bombard cells with DNA‐coated gold particles as described previously (Schiedlmeier et al., 1994). Plasmids carrying the artificial gene constructs were introduced into V.carteri nitA− strain 153‐48 by co‐transformation with plasmid pVcNR1 (Gruber et al., 1992; Schiedlmeier et al., 1994), containing the coding region of the V.carteri nitA gene plus downstream and upstream DNA. Bombarded cultures were cultivated in selective Volvox medium containing only nitrate as a nitrogen source.
Reverse transcription–PCR amplification and sequencing of chimeric transcripts
For reverse transcription–PCR the antisense primer 5′‐TTTGAGGCGCAATTCCG (pherophorin‐S) was used for transformants containing the β‐tubulin promoter–pherophorin‐S chimeric gene. The sense primer was 5′‐ATAACAAGCGACCACTAC (β‐tubulin). Products of PCR amplification were ligated into the SmaI site of pUC18 and sequenced.
Preparation of recombinant pherophorin‐S
Volvox transformants (clone PheroS‐T1), constitutively expressing the pherophorin‐S gene under the control of the β‐tubulin promoter, were grown in 20 l glass flasks under standard conditions. Spheroids were disrupted and centrifuged as described before. After centrifugation 33P‐labelled pherophorin‐S was added to facilitate identification. The supernatant was brought to 10% acetonitrile and passed over a C18 (octadecylsilane) cartridge (Millipore). The flow‐through was concentrated by lyophilization and applied to a preparative 8% SDS–PAGE gel. After autoradiography pherophorin‐S was eluted with water by diffusion, dialysed and lyophilized.
Preparation of the 50 kDa subtilisin fragment
Purified and 33P‐labelled pherophorin‐S was digested with 0.6 μg/μl subtilisin (Carlsberg, type VIII; Sigma) in 50 mM Tris–HCl, pH 8.0, 0.5% SDS for 1 h at 30°C and applied to 10% SDS–PAGE. After autoradiography the 50 kDa subtilisin fragment was eluted with water, dialysed and lyophilized. Deglycosylation of the 50 kDa subtilisin fragment was performed with anhydrous hydrogen fluoride at 0°C as described by Mort and Lamport (1977). The deglycosylated peptide was purified by reversed phase HPLC (SMART system; Pharmacia) on a 3 μm μRPC C2/C18 column (Pharmacia) by applying a 30 min linear gradient of 5–40% acetonitrile in 0.1% CF3CO2H at a flow rate of 200 μl/min.
Preparation of the phosphodiester of arabinose
Aliquots of 100 μg purified pherophorin‐S was mixed with 33P‐labelled pherophorin‐S and hydrolysed in 0.5 M trifluoracetic acid at 100°C for 2 h. The hydrolysate was dried in vacuo, redissolved in H2O and extracted twice with 1 vol. n‐butanol. The products were reduced with 0.5 M NaBH4 at 30°C for 45 min. After removal of H3BO3 (Laine et al., 1972) the sample was subjected to HPAEC using a CarboPac PA1 column (Dionex). A gradient of 0–0.5 M sodium acetate in 0.15 M NaOH was applied over 30 min at a flow rate of 1 ml/min. Monitoring of the eluate was performed by a pulsed amperometric detector (ED40; Dionex). Fractions of interest were passed through a column of AG 50W‐X8 H+ ion exchange resin, lyophilized and redissolved in 0.1 mM ammonium acetate containing 50% acetonitrile and subjected to electrospray mass spectrometry.
Molecular masses of fractions of interest were determined by electrospray mass spectrometry using the negative mode of a SSQ 7000 mass spectrometer (Finnigan).
Volvox spheroids suspended in 1 ml Volvox medium lacking glycerophosphate were pulse‐labelled with 0.4 mCi [14C]bicarbonate for 90 min. Pherophorin‐S was isolated as described above. The neutral sugar composition of pherophorin‐S and of the 50 kDa subtilisin fragment of pherophorin‐S was determined by radio gas chromatography of the alditol acetates as described by Wenzl and Sumper (1986a). For carbohydrate analysis of the phosphodiester, 14C‐labelled phosphodiester was isolated in the same way as described above. Complete hydrolysis of the phosphodiester was achieved in 6 M HCl at 100°C for 2 h. The liberated monosaccharides were identified by radio gas chromatography of alditol acetates. Unlabelled arabinose, xylose, mannose, galactose and glucose were added as internal standards.
Thin‐layer chromatography was performed on polyethyleneimine–cellulose plates (Schleicher & Schüll) according to Holst et al. (1989).
We wish to thank Dr R.Deutzmann and E.Hochmuth for mass spectrometry and for sequencing peptides. We also thank J.Nink for help with phosphodiester preparation and C.Friederich for expert technical assistance. This work was supported by the Deutsche Forschungsgemeinschaft (SFB 521).
- Copyright © 1997 European Molecular Biology Organization