Effect of impaired glycosylation on the biosynthesis of Semliki forest virus glycoproteins

Low glucose depletes glycan precursors, reduces site occupancy and galactosylation of a monoclonal antibody in CHO cell culture

Controlled feeding of glucose has been employed previously to enhance the productivity of recombinant glycoproteins but there is a concern that low concentrations of glucose could limit the synthesis of precursors of glycosylation. Here we investigate the effect of glucose depletion on the metabolism, productivity and glycosylation of a chimeric human-llama monoclonal antibody secreted by CHO cells. The cells were inoculated into media containing varying concentrations of glucose. Glucose depletion occurred in cultures with an initial glucose ≤5.5 mM and seeded at low density (2.5 × 10(5) cells/mL) or at high cell inoculum (≥2.5 × 10(6) cells/mL) at higher glucose concentration (up to 25 mM). Glucose-depleted cultures produced non-glycosylated Mabs (up to 51%), lower galactosylation index (GI <0.43) and decreased sialylation (by 85%) as measured by mass spectrometry and HPLC. At low glucose a reduced intracellular pool of nucleotides (0.03-0.23 fmoles/cell) was measured as well as a low adenylate energy charge (<0.57). Low glucose also reduced GDP-sugars (by 77%) and UDP-hexosamines (by 90%). The data indicate that under glucose deprivation, low levels of intracellular nucleotides and nucleotide sugars reduced the availability of the immediate precursors of glycosylation. These results are important when applied to the design of fed-batch cultures.

Differences in antigenicity of E2 in Semliki Forest virus particles and in infected cells

Using six monoclonal antibodies to epitopes a-f on the glycoprotein E2 of Semliki Forest virus (SFV) we found antigenic differences between E2 in infected cells and in virus particles, respectively, if glycosylation was impaired by 2-deoxy-D-glucose or inhibited by N-methyl-1-deoxynojirimycin. Furthermore we concluded that a conformational change of E2 takes place on virus budding.

Antiviral activity of some natural and synthetic sugar analogues

A number of natural and synthetic sugar analogues have been tested for their antiviral activity, using an influenza virus strain as a model. Hemagglutinating titres (HA) and cytopathic effect (CPE) were surveyed to estimate the virus production. It was found that introduction of the benzyl group into these sugars generally causes them to become antivirally active. Substitution with methyl, acetyl, uridyl and thiocyanyl groups or derivatization with azido, isopropylidene and benzylidene groups were without effect. All sugars containing the 2-deoxy-2-acetamido group were inactive.

The significance of carbohydrate trimming for the antigenicity of the Semliki Forest virus glycoprotein E2

Six groups, designated a-f, of noncompeting murine monoclonal antibodies to the envelope glycoprotein E2 of Semliki Forest virus (SFV) have been used to analyze antigenic changes caused by differences in the carbohydrate chain composition of the envelope glycoprotein E2 in the virion. Deletion of terminal sialic acids as observed in virus progeny from mosquito cells did not affect antigenic properties. Inhibition of the trimming pathway in infected chicken cells by the mannosidase I inhibitor dMM led to infectious virus particles containing mannose-rich oligosaccharides of the composition Man9(GlcNAc)2 in the envelope glycoproteins. This alteration had no effect on antigenicity. If inhibition was, however, performed with MdN which acts on alpha-glucosidase giving rise to virions with glycoproteins containing three additional glucose residues in the carbohydrate chains [Glc3Man7,8,9(GlcNAc)2], significant antigenic changes were observed. The six epitopes were differently affected by the underlying structural change and the pattern of exposition of epitopes was not identical with that observed after cleavage of intramolecular disulfide bonds. Concomitantly, the cleavage rate of gp62, the intracellular precursor molecule of the glycoproteins E2 and E3 of the virus particle, was reduced causing a reduction of virus yield. It is concluded that the existence of untrimmed carbohydrate chains is sufficient to allow SFV maturation. The trimming reactions improve this process in a matter suggesting that the carbohydrate chains influence intracellular traffic (addressing) of the respective glycoprotein.

Inhibition of the multiplication of enveloped and non-enveloped viruses by glucosamine

Glucosamine can inhibit the development of viral cytopathogenic effect and the production of infective viral particles of both enveloped and non-enveloped viruses. The extent of antiviral activity is dependent on drug concentration, composition of the culture medium and type of cell host.

Investigation of the role of glycans for the biological activity of Semliki Forest virus grown in Aedes albopictus cells using inhibitors of asparagine-linked oligosaccharides trimming

The effects of N-linked-oligosaccharide-processing inhibitors on the formation of Semliki Forest virus (SFV) in C6/36 Aedes albopictus cells were investigated. The glycosidase inhibitors deoxynojirimycin, deoxymannojirimycin and swainsonine prevented the formation of Endo-H resistant structures, but had little effect on virus formation and on the biological activities of the virus. Tunicamycin greatly inhibited virus formation, but had little effect on cell-cell fusion from within and the cleavage of p 62. These results indicate that correct glycosylation is not a prerequisite for biological activities of SFV, whereas glycosylation per se is needed for virus production.

Characterization of the N-linked high-mannose oligosaccharides of the insulin pro-receptor and mature insulin receptor subunits

The insulin receptor is synthesized as a 190,000-Mr single-chain precursor that contains exclusively asparagine-N-linked high-mannose-type carbohydrate chains. In this study we have characterized the structure of the pro-receptor oligosaccharides. IM-9 lymphocytes were pulse-chase-labelled with [3H]mannose, and the insulin pro-receptor was isolated by immunoprecipitation and SDS/polyacrylamide-gel electrophoresis. The pro-receptor oligosaccharides were removed from the protein backbone with endoglycosidase H and analysed by h.p.l.c. Immediately after a [3H]mannose pulse the largest oligosaccharide found in the pro-receptor was Glc1Man9GlcNAc2; this structure represented only a small fraction (3%) of the total. The predominant oligosaccharides present in the pro-receptor were Man9GlcNAc2 (25%) and Man8GlcNAc2 (48%). Smaller oligosaccharides were also detected: Man7GlcNAc2 (18%), Man6GlcNAc2 (3%) and Man5GlcNAc2 (3%). The relative distribution of the different oligosaccharides did not change at 1, 2 or 3 h after the pulse with the exception of the rapid disappearance of the Glc1Man9GlcNAc2 component. The mature alpha- and beta-subunits of the insulin receptor are known to contain both high-mannose-type and complex-type oligosaccharides. We have also examined here the structure of the high-mannose chains of these subunits. The predominant species in the alpha-subunit was Man8GlcNAc2 whereas in the beta-subunit it was Man7GlcNAc2. These results demonstrate that most (approx. 75%) oligosaccharides of the insulin pro-receptor are chains of the type Man8GlcNAc2 or Man9GlcNAc2. Thus, assuming that a Glc3Man9GlcNAc2 species is transferred co-translationally, carbohydrate processing of the pro-receptor appears to be very rapid and limited to the removal of the three glucose residues and one mannose residue. Further mannose removal does not occur until the pro-receptor has been proteolytically cleaved. In addition, the degree of mannose trimming appears to be different in the alpha- and beta-subunits.

Role of 2-deoxy-D-glucose in the inhibition of phagocytosis by mouse peritoneal macrophage

2-Deoxy-D-glucose inhibits Fc and complement receptor-mediated phagocytosis of mouse peritoneal macrophages. To understand the mechanism of this inhibition, we analyzed the 2-deoxy-D-glucose metabolites in macrophages under phagocytosis inhibition conditions and conditions of phagocytosis reversal caused by glucose, mannose and 5-thio-D-glucose, and compared their accumulations under these conditions. Macrophages metabolized 2-deoxy-D-glucose to form 2-deoxy-D-glucose 6-phosphate, 2-deoxy-D-glucose 1-phosphate, UDP-2-deoxy-D-glucose, 2-deoxy-D-glucose 1, 6-diphosphate, 2-deoxy-D-gluconic acid and 2-deoxy-6-phospho-D-gluconic acid. The level of bulk accumulation as well as the accumulation of any of these 2-deoxy-D-glucose metabolites did not correlate with changes in macrophage phagocytosis capacities caused by the reversing sugars. 2-Deoxy-D-glucose inhibited glycosylation of thioglycolate-elicited macrophage by 70-80%. This inhibition did not cause phagocytosis inhibition, since (1) the reversal of phagocytosis by 5-thio-D-glucose was not followed by increases in the incorporation of radiolabelled galactose, glucosamine, N-acetylgalactosamine or fucose; (2) cycloheximide at a concentration that inhibited glycosylation by 70-80% did not affect macrophage phagocytosis. The inhibition of protein synthesis by 2-deoxy-D-glucose similarly could not account for phagocytosis inhibition, since cycloheximide, when used at a concentration that inhibited protein synthesis by 95%, did not affect phagocytosis. 2-Deoxy-D-glucose lowered cellular nucleoside triphosphates by 70-99%, but their intracellular levels in the presence of different reversing sugars did not correlate with the magnitude of phagocytosis reversal caused by these sugars. The results show that 2-deoxy-D-glucose inhibits phagocytosis by a mechanism distinct from its usual action of inhibiting glycosylation, protein synthesis and depleting energy supplies, mechanisms by which 2-deoxy-D-glucose inhibits other cellular processes.

Detection of immunologically cross-reacting capsid protein of alphaviruses on the surfaces of infected L929 cells

Hyperimmune, but not normal immune, monospecific antiserum made to capsid protein of Sindbis virus (SIN) was found to cause cytolysis equally well of both SIN- and Semliki Forest virus-infected L929 cells in antibody-dependent, complement-mediated cytotoxicity assays. The cell surface reactivity of the hyperimmune antiserum was also demonstrated by solid-phase radioimmune assays with unfixed infected cells or infected cells fixed with low concentrations of glutaraldehyde (0.025%) before reactivity with antisera. Higher concentrations of glutaraldehyde lowered the sensitivity of detection. Purified SIN capsid protein specifically inhibited antibody-dependent, complement-mediated cytotoxicity by the monospecific anti-capsid protein serum on SIN- and Semliki Forest virus-infected target cells. That hyperimmune anti-SIN serum also cross-reacts with capsid protein on the surface of Semliki Forest virus-infected cells was suggested by the fact that capsid protein inhibited cross-cytolysis in the antibody-dependent, complement-mediated cytotoxicity assay. The latter antiserum was collected after repeated injections of purified virions over a 9-month period. The results suggest that hyperimmune monospecific antisera made to SIN capsid protein or hyperimmune antisera to SIN or Semliki Forest virions detect homologous and cross-reacting capsid protein determinants on the surface of infected cells.

Glucose requirement for induction by sodium butyrate of the glycoprotein hormone alpha subunit in HeLa cells

Butyric acid produces multiple effects on mammalian cells in culture, including alterations in morphology, depression of growth rate, increased histone acetylation, and modified production of various proteins and enzymes. The latter effect is exemplified by the induction in HeLa cells of the glycoprotein hormone alpha subunit by millimolar concentrations of the fatty acid. This report demonstrates that increased subunit accumulation in response to sodium butyrate is strikingly dependent on the presence of glucose (or mannose) in the growth medium. In contrast, basal levels of subunit synthesis are only marginally affected when the culture medium is supplemented with one of a variety of hexoses. An increase in the accumulation of HeLa alpha does not occur in medium containing pyruvate as the energy source, and sustained induction requires the simultaneous and continued presence of both glucose and butyrate. The effects of butyrate on HeLa cell morphology and subunit induction can be separated, since the latter is glucose-dependent while the former is not. Failure of butyrate to induce alpha in medium containing pyruvate does not result from restricted subunit secretion, since the levels of intracellular alpha are not increased disproportionately relative to those in the medium. The hexoses which support induction of HeLa alpha (glucose greater than or equal to mannose greater than galactose greater than fructose) are identical to those which have been shown previously to stimulate the glucosylation of lipid-linked oligosaccharides and enhance the synthesis of certain glycoproteins. Labeling of various glycosylation intermediates with [3H]mannose indicates that in glucose medium there is a decrease in the level of radioactivity associated with both dolicholpyrophosphoryl oligosaccharide and cellular glycoproteins and a concomitant increase in the fraction of label recovered in secreted glycoproteins. Butyrate also causes a decrease in [3H]mannose-labeled cellular glycoproteins and an increase in tritiated extracellular glycoproteins, particularly in glucose medium. Likewise, glucose stimulates the incorporation of [3H]glucosamine into immunoprecipitable alpha subunit relative to the bulk of HeLa-secreted glycoproteins, and this is further enhanced by butyrate. However, as demonstrated by lectin chromatography of conditioned media, a nonglycosylated subunit does not accumulate in pyruvate medium, either in the absence or presence of butyrate.(ABSTRACT TRUNCATED AT 400 WORDS)

Identification of immunologically cross-reactive proteins of Sindbis virus: evidence for unique conformation of E1 glycoprotein from infected cells

Hyperimmune antisera to purified Sindbis (SIN) or Semliki Forest (SF) virus were used to identify alphavirus-specific and cross-reactive proteins in virions and infected cells. The hyperimmune sera participated in homologous and cross-cytolysis of alphavirus-infected cells, and the use of monospecific antisera to SIN structural proteins suggested that E1 and E2 could serve as target proteins in cytolysis. Proteins from purified virions or infected cells were extracted with Nonidet P-40, denatured by procedures for sodium dodecyl sulfate-polyacrylamide gel electrophoresis, transferred to nitrocellulose solid supports, and reacted with hyperimmune sera and 125I-labeled protein A (immunoblotting on denatured proteins). Alternatively, native proteins extracted by mild Nonidet P-40 treatment were precipitated with hyperimmune sera before denaturation by sodium dodecyl sulfate-polyacrylamide gel electrophoresis. After immunoblotting, homologous antiserum reacted with the virus structural proteins E1, E2, capsid extracted from purified virions, and the counterparts of these proteins extracted from infected cells. In addition, PE2 and a 92,000-molecular-weight protein from infected cells reacted with homologous antiserum. These proteins were also immunoprecipitated with homologous antiserum. After immunoblotting, the Sindbis capsid protein was shown to be cross-reactive whether derived from purified virions or from infected cells; no cross-reactivity was observed with PE2 or E2 from either source, and the E1 glycoprotein was shown to be cross-reactive only when obtained from virions. However, the E1 glycoprotein could be cross-immunoprecipitated from infected cells (as well as from disrupted virions), and, in addition, capsid and a 92,000-molecular-weight protein were cross-immunoprecipitated from infected cells. These results suggest that a native conformation of the cell-associated E1 glycoproteins may be required for immunological cross-reactivity (immune precipitation), whereas virion but not cell-associated E1 retains immunological cross-reactivity after denaturation (immunoblot technique). The findings extend our previously published evidence which suggested that alphavirus maturation is accompanied by a change in immunological cross-reactivity with respect to E1.

The biosynthetic pathway of the asparagine-linked oligosaccharides of glycoproteins

This review deals with the structure and addition of the different types of oligosaccharides to asparagine residues in proteins. This process occurs in several steps, first an oligosaccharide which contains N-acetylglucosamine mannose and glucose is built up joined to dolichyl diphosphate. The oligosaccharide is then transferred to a polypeptide chain, loses its glucose, and is modified by removal of some monosaccharides and addition of others giving rise to a variety of saccharides.

Cell interactions in the bone marrow microenvironment: role of endogenous colony-stimulating activity

Adherent stromal cells from mouse bone marrow inhibited the formation of granulocyte/monocyte (G/M) colonies induced in vitro by colony-stimulating factor (CSF). This inhibition occurred both when crude conditioned media obtained from various sources were used to induce colony formation or when a pure CSF preparation from mouse lung origin was tested. The inhibition did not appear to be toxic in nature since despite the lack of colony formation, progenitor CFU-C proliferated in the presence of stromal cells. Medium conditioned by adherent stromal cells was devoid of inhibitory activity when incorporated into the culture medium used for G/M colony formation, indicating that the inhibitory activity may not be present in a soluble form. Inhibitors of prostaglandins did not affect G/M colony formation. In contrast, D-glucose and a number of other free monosaccharides but not pyruvate lactate or glycerol induced formation of myeloid colonies in the presence of stromal cells. This did not require addition of exogenous CSF. Released factors concentrated from serum-free medium conditioned by stromal cells exhibited colony-stimulating activity provided that the medium contained a high glucose concentration during incubation. It is proposed that stromal cells produce a resident CSF that, in contrast to exogenous CSF species, is capable of inducing myelopoiesis within the bone and marrow stroma.

The relationship between glycosylation and glycoprotein metabolism of mouse neuroblastoma N18 cells

Two inhibitors of glycosylation, glucosamine and tunicamycin, were utilized to examine the effect of glycosylation inhibition in mouse neuroblastoma N18 cells on the degradation of membrane glycoproteins synthesized before addition of the inhibitor. Treatment with 10 mM-glucosamine resulted in inhibition of glycosylation after 2h, as measured by [3H]fucose incorporation into acid-insoluble macromolecules, and in a decreased rate of glycoprotein degradation. However, these results were difficult to interpret since glucosamine also significantly inhibited protein synthesis, which in itself could cause the alteration in glycoprotein degradation [Hudson & Johnson (1977) Biochim. Biophys. Acta 497, 567-577]. N18 cells treated with 5 microgram of tunicamycin/ml, a more specific inhibitor of glycosylation, showed a small decrease in protein synthesis relative to its effect on glycosylation, which was inhibited by 85%. Tunicamycin-treated cells also showed a marked decrease in glycoprotein degradation in experiments with intact cells. The inhibition of glycoprotein degradation by tunicamycin was shown to be independent of alterations in cyclic AMP concentration. Polyacrylamide-gel electrophoresis of isolated membranes from N18 cells, double-labelled with [14C]fucose and [3H]fucose, revealed heterogeneous turnover rates for specific plasma-membrane glycoproteins. Comparisons of polyacrylamide gels of isolated plasma membranes from [3H]fucose-labelled control cells and [14C]fucose-labelled tunicamycin-treated cells revealed that both rapidly and slowly metabolized, although not all, membrane glycoproteins became resistant to degradation after glycosylation inhibition.

Restricted replication of two alphaviruses in ricin-resistant mouse L cells with altered glycosyltransferase activities

Two mouse L cell variant lines (CL 3 and CL 6) selected for resistance to the toxic plant lectin ricin were restricted in their ability to replicate the two alphaviruses Sindbis virus and Semliki Forest virus. CL 3 cells have been shown to exhibit increased CMP-sialic acid:glycoprotein sialyltransferase and GM3 synthetase activities, whereas CL 6 cells have been shown to contain decreased UDPgalactose:glycoprotein galactosyltransferase and UDP-N-acetylglucosamine:glycoprotein N-acetylglucosaminyltransferase activities. The adsorption of Sindbis virus to CL 6 cells was considerably reduced, suggesting that the loss or inaccessibility of the receptors for Sindbis virus accounted for a major defect in virus production in these cells. In contrast, CL 3 synthesized Sindbis viral RNA and proteins but were unable to convert the precursor glycoprotein PE2 to the structural protein E2. The cleavage of PE2 to E2 was also blocked in both CL 3 and CL 6 cells infected with Semliki Forest virus.

Immediate glycosylation of Sindbis virus membrane proteins

The mechanism by which the membrane proteins of Sindbis virus are initially glycosylated during growth of the virus in chick cells was studied. The experiments suggest strongly that the two viral glycoproteins are glycosylated before release from the polysome, and that this glycosylation involves transfer of a large 1800 dalton oligosaccharide to the polypeptide chains. The donor of the oligosaccharide is most probably a lipid.

Rabies virus protein synthesis in infected BHK-21 cells

Rabies virus specific polypeptide synthesis was examined under hypertonic conditions, which selectively inhibit cellular protein synthesis. The rabies virus proteins (L, G, N, M1, M2) were synthesized throughout the course of infection, with little change in their relative rates of synthesis. The rates of synthesis of the G and M1 polypeptides were more sensitive to increasing osmolarity than those of the L, N, and M2 polypeptides. Extrapolation to isotonicity of the results obtained under hypertonic conditions indicated that the molar ratios of the polypeptides synthesized under normal conditions were 0.4 (L), 64 (G), 100 (N), 75 (M1) and 35 (M2). A high-molecular-weight polypeptide (190,000), designated polypeptide L, was repeatedly detected both in infected cells and in extracellular virus. The estimated number of L polypeptide molecules per virion was 33. The synthesis of a viral glycoprotein precursor, designated gp78, , preceded the appearance of the mature viral glycoprotein in infected cells labeled with [3H]glucosamine under isotonic conditions. In cells labeled under hypertonic conditions, little or no mature viral glycoprotein was detected, but a virus-specific glycoprotein with an electrophoretic mobility similar to that of gp78 was observed. This glycoprotein could be chased into mature viral glycoprotein when the hypertonic conditions were made isotonic. These results suggest that a reversible block of viral glycoprotein synthesis occurs under hypertonic conditions.

Interference of nucleoside diphosphate derivatives of 2-deoxy-D-glucose with the glycosylation of virus-specific glycoproteins in vivo

The predominant effect of 2-deoxy-D-glucose on chick embryo cells infected with Semliki Forest virus is an interference with glycosylation of virus-specific glycoproteins; this results in a block of synthesis of infectious virus. Incorporation of radioactive mannose is blocked severely in the presence of 2-deoxyglucose in the cultural medium although it is readily phosphorylated and subsequently activated by GTP to yield GDP-mannose, which accumulates under these conditions. The intracellular concentrations of GDP-mannose and UDP-N-acetyl-D-hexosamine are not reduced in the presence of the inhibitor. An equimolar concentration of mannose in the cultural medium competes with the inhibitory effect of the deoxysugar and drops the cellular pool of GDP-2-deoxy-D-glucose below the level of detection, at the same time restoring the synthesis of infectious virus. When the intracellular concentration of UDP-2-deoxyglucose is reduced by addition of glucose into the cultural medium the inhibition of virus synthesis by the deoxysugar and the concentration of GDP-2-deoxyglucose within the cells remain near to the values when the inhibitor is present alone. It is concluded that among the metabolites of 2-deoxyglucose which occur in vivo after addition of 2-deoxyglucose to the culture medium, GDP-2-deoxyglucose is the agent responsible for inhibition of glycosylation of viral glycoproteins.

Suppression of glycoprotein formation of Semliki Forest, influenza, and avian sarcoma virus by tunicamycin

Tunicamycin, a new antibiotic, halts the formation of physical particles of Semliki forest and fowl plague virus, whereas avian oncornavirus particles which show a reduction in infectivity and do not contain detectable labeled glycoprotein are released in the presence of the drug. In Semliki forest virus-infected cells only the protein moieties of the glycoproteins could be labeled. In cells infected with fowl plague and avian sarcoma virus neither intact glycoproteins nor their protein moieties could be detected. By using a protease inhibitor (N-alpha-p-tosyl-L-lysin chloromethyl ketone, TLCK) it could be shown, however, that the carbohydrate-free hemagglutinin precursor of influenza virus is synthesized but is presumably degraded by intracellular proteases in the absence of TLCK as a consequence of the lack of carbohydrate.

Early synthesis of Semliki Forest virus-specific proteins in infected chicken cells

Cells preinfected with fowl plague virus followed by treatment with actinomycin D are a suitable system for studying early protein synthesis in cells infected with Semliki forest virus. One and one-half hours after superinfection, three new nonstructural proteins (NVP) were detected: NVP 145, NVP, 112, and NVP 65. They appeared in parallel with a low incorporation of mannose at the beginning of the infectious cycle. Behavior on chasing suggested a precursor relationship of NVP 112 to the envelope glycoproteins. Two kinds of NVP 65 are described, both of which are varieties of NVP 68 with an incomplete mannose content. One type, detected early after infection, was converted into NVP 68 by supplementary glycosylation. The second, late type was stable. It contains fucose and resembles the NVP 65 observed after impairment of glycosylation. The mechanism of NVP 68 glycosylation is discussed. The presence of the complete carbohydrate moiety is crucial for the cleavage of NVP 68 into the envelope proteins E2 and E3 and, thus, for virus maturation. Only the complete form of NVP 68 was precipitated by envelope-specific antisera. A large production of NVP 78 is a further feature of the early events in infected cells. It is not related to the structural proteins.

Effects of 2-deoxyglucose, glucosamine, and mannose on cell fusion and the glycoproteins of herpes simplex virus

2-Deoxyglucose and glucosamine were found to inhibit cell fusion caused by a syncytial mutant of herpes simplex virus and to inhibit the glycosylation of viral glycoproteins in the infected cells. The inhibition of fusion and the inhibition of glycosylation caused by 2-deoxyglucose were substantially prevented when mannose was also present during infection. When glycosylation was inhibited, three new bands were found in major glycoprotein region on sodium dodecyl sulfate-polyacrylamide gels. These bands may be precursors to the normal glycoproteins. The correlation between fusion and glycosylation in the presence of 2-deoxyglucose, glucosamine, and mannose suggests that the cells cannot fuse if their glycoproteins have a considerably reduced carbohydrate content.